ISO 19144-1:2009
(Main)Geographic information — Classification systems — Part 1: Classification system structure
Geographic information — Classification systems — Part 1: Classification system structure
ISO 19144-1:2009 establishes the structure of a geographic information classification system, together with the mechanism for defining and registering the classifiers for such a system. It specifies the use of discrete coverages to represent the result of applying the classification system to a particular area and defines the technical structure of a register of classifiers in accordance with ISO 19135.
Information géographique — Systèmes de classification — Partie 1: Structure de système de classification
L'ISO 19144-1:2009 établit la structure d'un système de classification d'informations géographiques, ainsi que le mécanisme de définition et d'enregistrement des classificateurs pour un tel système. Elle spécifie l'utilisation de couvertures discrètes pour représenter le résultat de l'application du système de classification à une zone particulière et définit la structure technique d'un registre de classificateurs, conformément à l'ISO 19135.
Geografske informacije - Klasifikacijski sistemi - 1. del: Struktura klasifikacijskega sistema
General Information
Relations
Standards Content (Sample)
SLOVENSKI STANDARD
01-december-2009
Geografske informacije - Klasifikacijski sistemi - 1. del: Struktura klasifikacijskega
sistema
Geographic information - Classification systems - Part 1: Classification system structure
Information géographique - Systèmes de classification - Partie 1: Structure de système
de classification
Ta slovenski standard je istoveten z: ISO 19144-1:2009
ICS:
35.240.70 Uporabniške rešitve IT v IT applications in science
znanosti
2003-01.Slovenski inštitut za standardizacijo. Razmnoževanje celote ali delov tega standarda ni dovoljeno.
INTERNATIONAL ISO
STANDARD 19144-1
First edition
2009-08-15
Geographic information — Classification
systems —
Part 1:
Classification system structure
Information géographique — Systèmes de classification —
Partie 1: Structure de système de classification
Reference number
©
ISO 2009
PDF disclaimer
This PDF file may contain embedded typefaces. In accordance with Adobe's licensing policy, this file may be printed or viewed but
shall not be edited unless the typefaces which are embedded are licensed to and installed on the computer performing the editing. In
downloading this file, parties accept therein the responsibility of not infringing Adobe's licensing policy. The ISO Central Secretariat
accepts no liability in this area.
Adobe is a trademark of Adobe Systems Incorporated.
Details of the software products used to create this PDF file can be found in the General Info relative to the file; the PDF-creation
parameters were optimized for printing. Every care has been taken to ensure that the file is suitable for use by ISO member bodies. In
the unlikely event that a problem relating to it is found, please inform the Central Secretariat at the address given below.
© ISO 2009
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying and microfilm, without permission in writing from either ISO at the address below or
ISO's member body in the country of the requester.
ISO copyright office
Case postale 56 • CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax + 41 22 749 09 47
E-mail copyright@iso.org
Web www.iso.org
Published in Switzerland
ii © ISO 2009 – All rights reserved
Contents Page
Foreword .iv
Introduction.v
1 Scope.1
2 Conformance .1
2.1 Classes .1
2.2 Conformance of a classification system.1
2.3 Conformance of a register of classifiers.1
2.4 Representation of classification results .1
3 Normative references.1
4 Terms, definitions and abbreviated terms.2
4.1 Terms and definitions .2
4.2 Abbreviated terms.4
5 Classification systems.5
5.1 Concept .5
5.2 Classification and legend .7
5.3 Hierarchical versus non-hierarchical systems.8
5.4 A priori and a posteriori classification systems .8
5.5 Structure of classified data .9
5.6 A classification data set.13
6 Management of classifiers .14
6.1 General .14
6.2 Concept dictionary register for a classification scheme .15
6.3 Management of classifiers through registration.15
6.4 Register structure.15
Annex A (normative) Abstract test suite .24
Annex B (informative) A priori and a posteriori classification systems .26
Bibliography.30
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards bodies
(ISO member bodies). The work of preparing International Standards is normally carried out through ISO
technical committees. Each member body interested in a subject for which a technical committee has been
established has the right to be represented on that committee. International organizations, governmental and
non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the
International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization.
International Standards are drafted in accordance with the rules given in the ISO/IEC Directives, Part 2.
The main task of technical committees is to prepare International Standards. Draft International Standards
adopted by the technical committees are circulated to the member bodies for voting. Publication as an
International Standard requires approval by at least 75 % of the member bodies casting a vote.
Attention is drawn to the possibility that some of the elements of this document may be the subject of patent
rights. ISO shall not be held responsible for identifying any or all such patent rights.
ISO 19144-1 was prepared jointly by the Food and Agriculture Organization (FAO) of the United Nations and
Technical Committee ISO/TC 211, Geographic information/Geomatics under a cooperative agreement
between the two organizations.
ISO 19144 consists of the following parts, under the general title Geographic information — Classification
systems:
⎯ Part 1: Classification system structure
The following part is under preparation:
⎯ Part 2: Land cover classification system (LCCS)
iv © ISO 2009 – All rights reserved
Introduction
This part of ISO 19144 is based on publications of the Food and Agriculture Organization (FAO) of the United
[1][2]
Nations . The first in a series of International Standards related to geographic classification systems, it
defines the structure of such systems, together with the mechanism for defining and registering classifiers.
Since there are many different possible application areas, there is no single classification system that will
serve all needs. The method by which classifiers are defined depends upon the application area. In addition,
the classifiers used within a particular application area might not be adequate for all situations encountered
within that application area and could need to be augmented over time. To facilitate extension of the set of
classifiers in a particular application area, classifiers are registered in a register structure compliant with
ISO 19135. This allows the set of classifiers to be maintained. The use of the ISO 19135 registration
mechanism allows for separate registers to be defined for different sets of classifiers within multiple
information communities, thereby satisfying application needs. This approach allows for independence
between information communities, but also allows relationships to be developed between different
classification systems that potentially allow the conversion, or partial conversion, of data from one
classification system to another, or the fusion of data from two separate sources.
The concept of classification systems is well known in the geographic information community. A classification
system can be used to subdivide any geographic area into small units, each of which carries an identifier that
describes its type. The results can then be represented as a discrete coverage as described in ISO 19123.
Many such classification systems can be defined to address any geographic area. Different application areas
and different information communities can define their own classification systems. However, if the
classification system is defined in a compatible way, interaction between different information communities
becomes possible. In addition, in a particular application area, it is desirable that there be a few
well-established classification systems, and that these themselves be standardized within information
communities.
This part of ISO 19144 describes the common structure, while subsequent parts will allow for the
standardization of specific classification systems.
A coverage is a function that returns values from its range for any direct position within its spatial, temporal or
spatiotemporal domain. A discrete coverage is a function that returns the same feature attribute values for
every direct position within any single spatial object, temporal object or spatiotemporal object in its domain.
The domain is an area covered by the coverage function, and the discrete coverage breaks that area down
into a set of spatial, temporal or spatiotemporal objects. The geometry of the discrete coverage used to
represent the results of applying a classification system can be any type of discrete coverage — for example,
a set of polygons fitted together like a jig-saw puzzle, a set of grid cells, or a set of points or curves.
A classification system consists of a set of classifiers. These classifiers may be algorithmically defined, or
established according to a set of classification system definitions. The classifiers are
application-area-dependent and are or will be defined in the other parts of ISO 19144 or other standards or
publications. A register allows for the maintenance of a set of classifiers for a particular application area. A
spatial, temporal or spatiotemporal object defined in terms of a set of classifiers is a classified object.
There is a commonality between conventional geographic features and classified objects. A feature is defined
in ISO 19101 as an abstraction of real world phenomena. An example of a class of feature is a building, and a
particular building, e.g. the UN building in New York, is an instance of a feature class. Conventional
geographic features are atomic units that are assembled to build one type of geographic information data set.
A classification system works in the opposite manner, from the top down, by successively decomposing the
whole within a coverage area. Classified objects are features, in that they are an abstraction of a real world
phenomena, but classified objects are not atomic, because they are necessarily related to each other by the
classifiers that decompose the whole. In a simple example of a classification system, the earth as a whole can
be covered by either “land” or “water”, and two classifiers can be defined partitioning the attribute range into
two, identifying objects as being either land or water. Any particular area on the earth, corresponding to a
classified object, would be of type “land” or “water”.
ISO 19135 specifies that a technical standard be required to define the item classes in any conformant
register. This part of ISO 19144 defines schemas for registers conformant to ISO 19135 and serves as the
technical standard that defines the item classes required for the registration of classifiers. It establishes a set
of rules for specifying definitions that can be used in a particular context to establish classified objects.
Registers of classifiers can serve as sources of reference for similar registers established by other geographic
information communities as part of a system of cross-referencing. Cross-referencing between respective items
in registers of classifiers might be difficult in cases where the structure of registers differs between information
communities. This part of ISO 19144 can serve as a guide for different information communities for the
development of compatible registers that can support a system of classifier cross-referencing.
The structure of a classification system together with the mechanism of defining and registering classifiers
defined in this part of ISO 19144 is general and can be applied to many different information-community-
defined classification systems, including soil, landform, vegetation, urbanization and systems for
understanding biodiversity and climate change. The use of this document will allow the relationship between
different classification systems to be described.
vi © ISO 2009 – All rights reserved
INTERNATIONAL STANDARD ISO 19144-1:2009(E)
Geographic information — Classification systems —
Part 1:
Classification system structure
1 Scope
This part of ISO 19144 establishes the structure of a geographic information classification system, together
with the mechanism for defining and registering the classifiers for such a system. It specifies the use of
discrete coverages to represent the result of applying the classification system to a particular area and defines
the technical structure of a register of classifiers in accordance with ISO 19135.
The structure can be used to develop specific classification systems that address particular application areas,
specified in other parts of ISO 19144.
2 Conformance
2.1 Classes
Three conformance classes are identified in this part of ISO 19144.
2.2 Conformance of a classification system
Any classification system for which conformance to this part of ISO 19144 is claimed shall be in accordance
with Annex A (see A.2).
2.3 Conformance of a register of classifiers
Any register of classifiers for which conformance to this part of ISO 19144 is claimed shall be in accordance
with Annex A (see A.3) and ISO 19135:2005, A.1.
2.4 Representation of classification results
Any legend of classifiers for which conformance to this part of ISO 19144 is claimed shall be in accordance
with Annex A (see A.4).
3 Normative references
The following referenced documents are indispensable for the application of this document. For dated
references, only the edition cited applies. For undated references, the latest edition of the referenced
document (including any amendments) applies.
ISO/TS 19103:2005, Geographic information — Conceptual schema language
ISO 19110:2005, Geographic information — Methodology for feature cataloguing
ISO 19115, Geographic information — Metadata
ISO 19123, Geographic information — Schema for coverage geometry and functions
ISO 19135:2005, Geographic information — Procedures for item registration
4 Terms, definitions and abbreviated terms
For the purposes of this document, the following terms, definitions and abbreviated terms apply.
4.1 Terms and definitions
4.1.1
a posteriori classification
classification (4.1.4) scheme based upon definition of classes after clustering the field samples collected
NOTE 1 Taken from FAO LCCS version 2 (see Reference [2]).
NOTE 2 See Annex B for an examination of this and a priori classification (4.1.2).
4.1.2
a priori classification
classification (4.1.4) scheme structured so that the classes are abstract conceptualizations of the types
actually occurring
NOTE 1 Taken from FAO LCCS version 2 (see Reference [2]).
NOTE 2 The approach is based upon the definition of classes before any data collection actually takes place.
NOTE 3 See Annex B for an examination of this and a posteriori classification (4.1.1).
4.1.3
classified object
spatial object, temporal object or spatiotemporal object assigned to a specific legend class (4.1.16)
4.1.4
classification
abstract representation of real world phenomena using classifiers (4.1.6)
4.1.5
classification system
system for assigning objects to classes
4.1.6
classifier
definition used to assign objects to legend classes (4.1.16)
NOTE Classifiers can be defined algorithmically or according to a set of classification system (4.1.5) specific rules.
4.1.7
coverage
feature (4.1.10) that acts as a function to return values from its range (4.1.17) for any direct position within its
spatial, temporal or spatiotemporal domain (4.1.9)
[ISO 19123:2005]
EXAMPLE Raster image, polygon overlay, digital elevation matrix.
NOTE A coverage is a feature that has multiple values for each attribute type, where each direct position within the
geometric representation of the feature has a single value for each attribute type.
2 © ISO 2009 – All rights reserved
4.1.8
discrete coverage
coverage (4.1.7) that returns the same feature attribute (4.1.11) values for every direct position within any
single spatial object, temporal object, or spatiotemporal object in its domain (4.1.9)
[ISO 19123:2005]
NOTE The domain of a discrete coverage consists of a finite set of spatial, temporal, or spatiotemporal objects.
4.1.9
domain
well-defined set
[ISO/TS 19103:2005]
NOTE Domains are used to define the domain and range (4.1.17) of operators and functions.
4.1.10
feature
abstraction of real world phenomena
[ISO 19101:2002]
EXAMPLE The phenomenon “Eiffel Tower” can be classified with other similar phenomena into a feature type
“tower”.
NOTE A feature can occur as a type or an instance. In this part of ISO 19144, type is meant unless otherwise
specified.
4.1.11
feature attribute
characteristic of a feature (4.1.10)
[ISO 19101:2002]
4.1.12
feature concept dictionary
dictionary that contains definitions of, and related descriptive information about concepts that may be specified
in detail in a feature (4.1.10) catalogue
1)
[ISO 19126:— ]
4.1.13
identifier
linguistically independent sequence of characters capable of uniquely and permanently identifying that with
which it is associated
[ISO 19135:2005]
4.1.14
item class
set of items with common properties
[ISO 19135:2005]
NOTE “Class” is used in this context to refer to a set of instances, not the concept abstracted from that set of
instances.
1) To be published.
4.1.15
legend
application of a classification (4.1.4) in a specific area using a defined mapping scale and specific data set
NOTE Taken from FAO LCCS version 2 (see Reference [2]).
4.1.16
legend class
class resultant from the application of a classification (4.1.4) process
NOTE The result of a classification process is termed legend class in this part of ISO 19144 in order to avoid
confusion with the term “class” as used in UML modelling.
4.1.17
range
〈coverage〉 set of feature attribute (4.1.11) values associated by a function with the elements of the
domain (4.1.9) of a coverage (4.1.7)
[ISO 19123:2005]
4.1.18
register
set of files containing identifiers (4.1.13) assigned to items with descriptions of the associated items
[ISO 19135:2005]
4.1.19
registry
information system on which a register (4.1.18) is maintained
[ISO 19135:2005]
4.1.20
technical standard
standard containing the definitions of item classes (4.1.14) requiring registration
[ISO 19135:2005]
4.1.21
vector geometry
representation of geometry through the use of constructive geometric primitives
[ISO 19107:2003]
4.2 Abbreviated terms
4.2.1 General
CRS Coordinate Reference System
LCCS Land Cover Classification System
UML Unified Modelling Language
4.2.2 Notation
The conceptual schema specified in this part of ISO 19144 is described using the Unified Modelling Language
(UML), following the guidance of ISO/TS 19103.
4 © ISO 2009 – All rights reserved
Several model elements used in this schema are defined in other standards in the ISO 19100 series. By
2)
convention within this suite of International Standards, names of UML classes , with the exception of basic
data type classes, include a two-letter prefix that identifies the International Standard and the UML package in
which the class is defined.
UML classes defined in this part of ISO 19144 have the two-letter prefix “CL”.
Table 1 lists the other International Standards and the packages in which UML classes used in this part of
ISO 19144 have been defined.
Table 1 — Sources of externally defined UML classes
International
Prefix Package
Standard
CV ISO 19123 Coverage core and discrete coverages
DS ISO 19115 Metadata application information
GF ISO 19109 General feature model
GM ISO 19107 Geometry root
MD ISO 19115 Metadata entity set information
MI ISO 19115-2 Metadata entity set imagery
RE ISO 19135 Procedures for registration
SC ISO 19111 Spatial referencing by coordinates
TM ISO 19108 Temporal objects
5 Classification systems
5.1 Concept
A discrete coverage returns the same feature attribute for every direct position within any single geometric
object in its spatiotemporal domain. The spatiotemporal domain consists of a set of geometric objects that
together form the coverage.
EXAMPLE 1 The discrete coverage of postal zones within a country: each zone has a different code and it is not
possible to interpolate between these codes. Nevertheless, there could be a high level relationship between the codes.
The small country of Monaco is divided into five quartiers: “Moneghetti”, “La Condamine”, “Fontvieille”, “Monaco-Ville” and
“Monte-Carlo”. These political jurisdictions completely cover the area of the country. The area of Monaco can be
represented as a discrete coverage with five spatial objects where each object has the geometry of a polygon. The
attribute value for each spatial object is the name of the political jurisdiction. See Figure 1.
2) There is potential confusion between the use of the term “class” as used in UML, “class” as used in a classification
scheme and “item class” as used in the procedures for registration. Class as used in a classification scheme is termed
legend class in this part of ISO 19144.
Moneghetti La Condamine
Monte-Carlo
Monaco
Fontvieille Monaco-Ville
Figure 1 — Example of discrete coverage with polygon geometry
The geometry of the spatial objects associated with a discrete coverage can also be grid cells within a grid
structure. Each of the grid cells may carry an attribute.
EXAMPLE 2 A discrete coverage with grid geometry of the same area as used in the previous example. See Figure 2.
The figure legend identifies the instances of the attribute values that actually exist in the data.
Monte-Carlo
Monaco
Moneghetti
Monaco-Ville
Fontvieille
La Condamine
Figure 2 — Example of discrete coverage with grid geometry
NOTE The discrete coverages illustrated by the above examples are simple because only one relatively simple
attribute has been used. However, in reality the attributes for each of the coverage spatial objects can be very complex. In
order to describe land cover it is necessary to integrate a large number of descriptive parameters related to soil, biology
and density into a comprehensive land cover classification system. Such a classification system is of course
application-area-dependent. An oceanographer will have a different classification system than that of a meteorologist.
Classification systems can vary widely in different application areas, but for similar application areas there needs to be
some commonality in order for data from different sources to be used together.
6 © ISO 2009 – All rights reserved
A classification system allows one to define classifiers in order to partition the attribute range of a discrete
coverage to establish classified objects.
5.2 Classification and legend
Classification is an abstract representation of real world phenomena (i.e. the situation in the field) using
classifiers. A classification is a systematic framework with the names of the classes and the definitions used to
distinguish them, and the relation between classes. Classification thus necessarily involves definition of class
boundaries that must be clear and based upon objective criteria.
A classification system shall be
⎯ scale independent, meaning that the classes at all levels of the system shall be applicable at any scale or
level of detail, and
⎯ source independent, implying that it is independent of the means used to collect information.
NOTE 1 Scale independence and source independence exist in the general case of a classification system. When the
system is applied to real data, i.e. when a legend is created, the scale and source limitations of the data can create
restrictions on the legend.
A legend is the application of a classification in a specific area using a defined mapping scale and specific
data set. Therefore, a legend may contain only a proportion, or subset, of all possible classes of the
classification.
A legend shall be
⎯ scale dependent, and
⎯ source dependent.
EXAMPLE 1 Scale dependence: some elements might be too small to be delineated independently when collected at
a particular scale.
EXAMPLE 2 Source dependence: elements collected from different sources, such as an aerial photograph and a
satellite false colour composite image, could be different.
A legend is a selection of a set of the classifiers from all of the possible classifiers within a classification
system. A legend could correspond to the classifiers that apply to a single data set or it could be applicable to
a number of related data sets.
NOTE 2 A legend is analogous to the set of features within a feature catalogue, which is a subset of all of the possible
features within a feature concept dictionary.
EXAMPLE 3 The derivation of legend classes from a classification system: when applying the reference classification
system to a given geographic area with a specific data type, only a subset of the classes from the reference system will
occur in the real data. This set of classes is the legend with which the mapping system will be carried out. The reference
system in Figure 3 is shown as a matrix with two parameters: the height of plants, ranging from 5 m to greater than 50 m,
and the percentage crown cover, ranging from less than 10 % to 100 %. The classes are numbered with a code number
from 1 to 100. In this specific case, the legend will have only four reference classification classes, numbers 33, 47, 78 and
83.
Reference Classification System
>50 m 30 m >5 m
12 3 4
100%
DEFINED
GEOGRAPHIC
47 AREA
50%
given scale and data type
97 98 99 100
10%
mapping
derived
units
legend
[2]
Figure 3 — Legend as application of classification in a particular area
5.3 Hierarchical versus non-hierarchical systems
There are two basic forms of classification system: hierarchical and non-hierarchical.
a) Hierarchical classification systems
⎯ Classifiers shall be defined so that all classes at a specific level of the hierarchy are mutually
exclusive.
⎯ Criteria used to define a classifier at one level of a hierarchical classification shall not be repeated at
another level (e.g. criteria used to define a classifier at a lower level shall not be duplicated to a
higher level of the hierarchy).
NOTE 1 Most classification systems are hierarchically structured. Such a classification offers greater consistency,
owing to its ability to accommodate different levels of information — starting with structured broad-level classes —
which allow further systematic subdivision into more detailed subclasses. At the higher levels of the classification
system, few diagnostic criteria are used; whereas at lower levels, the number of diagnostic criteria increases.
b) Non-hierarchical classification systems
Classifiers shall be defined so that all classes are mutually exclusive.
NOTE 2 Non-hierarchical classification systems apply to only a single level of information, e.g. a single scale or
type of information.
5.4 A priori and a posteriori classification systems
Two approaches are defined for establishing classification systems: a priori and a posteriori classifications.
The first is based on the establishment of a classification scheme before the collection of the data to which it is
applied, whereas the second is based on the establishment of classes based on the common properties
identified in the data collected.
8 © ISO 2009 – All rights reserved
NOTE Both approaches have applications. An a posteriori approach to classification provides additional flexibility that
allows the classification scheme to more closely match the data. However, it can make it more difficult to compare data
collected and classified separately. If standardized rules are used to establish classes in an a posteriori classification
system, then these rules can be used to generate broader sets of classes that encompass multiple separate
classifications. The a priori and a posteriori classification system approaches are discussed in Annex B.
5.5 Structure of classified data
5.5.1 Systematic arrangement
Classification is the process of systematic arrangement of objects into groups on the basis of their attributes.
The set of classes established in a classification system allows one to organize the attributes of geographic
features so that the attribute value for a feature instance is the name of a class from the classification system.
The groups can be directly defined or parametric rules can be used to generate the groups. A subset of the
set of groups in a specific area becomes a legend.
NOTE The concept of classification is very general and is widely used in many scientific and other disciplines.
Most geographic attributes are distributed spatially. That is, instances of geographic attributes normally have
different values at different spatial locations. Geographic information with discrete attributes can be
represented as a discrete coverage or with vector geometry describing a boundary-defined spatial data set
(vector data). Since the same real-world phenomenon may be represented using either coverage or vector
data, the systematic arrangement of objects into groups forming the classification can be applied to both
coverage and boundary-defined spatial vector data.
5.5.2 Representation using discrete coverages
The result of classifying an area is a legend that may be represented as a type of discrete coverage, as
specified in ISO 19123. The range of a discrete coverage is a set of feature attribute values represented as a
set of records with a common schema defined by the classification system. The domain consists of a finite
collection of classified objects together with their direct positions. The geometric object and its associated
record form a geometry value pair.
The class CV_Coverage (see ISO 19123) represents a feature type in accordance with the general feature
model from ISO 19109:2005, Figure 4. Class CV_Coverage has three attributes: the domainExtent, the
rangeType and the commonPointRule. In discrete coverages, there is a direct one-to-one relationship where
each CV_GeometryValuePair links to corresponding attributes. Attribute domainExtent describes the extent of
the domain coverage. The data type EX_Extent is defined in ISO 19115. Attribute rangeType describes the
structure and composition of the attribute data record. Attribute rangeType makes use of the data type
RecordType, which is defined in ISO/TS 19103. A RecordType is a metaclass that describes the structure of a
set of records. A RecordType consists of a list of attribute name/data type pairs. A simple list is the most
common form of rangeType. Attribute commonPointRule identifies the procedure to be used for evaluating the
CV_Coverage at a position that falls either on a boundary between geometric objects or within the boundaries
of two or more overlapping geometric objects, where the geometric objects are either CV_DomainObjects or
CV_ValueObjects. The data type CV_CommonPointRule is defined ISO 19123.
Associated with a CV_Coverage is a specification of the coordinate reference system (CRS) to which the
objects in the domain are referenced. The coordinate reference system is defined in ISO 19111.
Also associated with the CV_Coverage class are the CV_Domain and the CV_AttributeValues classes. The
CV_Coverage links the set of domain objects to the set of records containing the attribute values. A
CV_DomainObject can be any spatial or temporal or spatiotemporal object. The domain objects of a
CV_DiscreteSurfaceCoverage are constrained to be instances of GM_Surface. There is one instance of
CV_AttributeValues (i.e. one data record) for each instance of CV_DomainObject (i.e. each grid cell or
polygon area).
See Figure 4.
<>
GF_FeatureType
(from general feature model)
<>
<>
CV_Coverage
(from coverage core)
CoordinateReferenceSystem
+ domainExtent[1.*] : EX_Extent
0.*
+ rangeType : RecordType
+ commonPointRule : CV_CommonPointRule
+collection +collection
Range
Domain
1 +CRS 1.* +domainElement 0.* +rangeElement
<>
CV_AttributeValues
CV_DomainObject
SC_CRS (from coverage core) (from coverage core)
(from spatial referencing by coordinates)
+ values : Record
+spatialElement +temporalElement
SpatialComposition TemporalComposition
0.* 0.*
<>
TM_GeometricPrimitive
GM_Object (from temporal objects)
(from geometry root)
Figure 4 — Elements of a coverage
CV_DiscreteCoverage is the subclass that returns the same record of feature attribute values for any direct
position within a single CV_DomainObject in its domain. Each geometry value pair consists of a domain object
(e.g. a grid cell or a polygon) and a record of feature attribute values. See Figure 5. In discrete coverages
there is a direct one-to-one relationship where each CV_GeometryValuePair is composed of a
CV_DomainObject such as a GM_Object and a CV_AttributeValues value.
10 © ISO 2009 – All rights reserved
<>
CV_Coverage
(from coverage core)
+ domainExtent[1.*] : EX_Extent
+ rangeType : RecordType
+ commonPointRule : CV_CommonPointRule
CV_DiscreteCoverage
(from discrete coverages)
0.* +collection
CoverageFunction
1.* +element
CV_GeometryValuePair
(from coverage core)
+ geometry : CV_DomainObject
+ value : Record
Figure 5 — Elements of a discrete coverage
The type of discrete coverage is based on the type of geometric object in the spatial domain. ISO 19123:2005,
Clause 6, identifies five types of discrete coverage with different geometries:
⎯ a discrete point coverage, consisting of a set of independent points;
⎯ a grid point coverage, consisting of a set of grid points;
⎯ a discrete curve coverage, consisting of a set of curves;
⎯ a discrete surface coverage, consisting of a set of surfaces, typically GM_Surface objects (polygons) or
congruent rectangles (grid cells) or regular hexagons, but possibly TIN objects or Thiessen Polygon
objects;
⎯ a discrete solid coverage, consisting of a set of solid volumes.
A classification system can make use of any type of geometric object in its spatial domain. It is possible to
establish classifiers to identify the range values for any type of discrete coverage. Discrete coverages that
partition an area are addressed by discrete surface coverage.
5.5.3 Discrete surface coverage
A classification system that makes use of a coverage whose range partitions an area shall make use of the
CV_Discrete surface coverage in accordance with ISO 19123:2005, 6.8. The use of GM_Surface as the
geometry element is illustrated in Figure 6.
CoverageFunction CV_GeometryValuePair
CV_DiscreteCoverage (from coverage core)
+collection +element
(from discrete coverages)
0.* 1.*
+ geometry : CV_DomainObject
+ value : Record
CoverageFunction
CV_SurfaceValuePair
CV_DiscreteSurfaceCoverage
+collection +element
(from discrete coverages)
(from discrete coverages)
1.*
+ geometry : GM_Surface
Figure 6 — Discrete surface coverage
The surfaces that constitute the domain of a discrete surface coverage can be mutually exclusive and
exhaustively partition an area or can be independent sub-areas. The type of coverage geometry used often
depends upon the method by which data was collected. Natural phenomena tend to have irregular boundaries
and are often represented as a set of congruent polygons. Some phenomena, types of sensor data or the
results of calculations tend to be organized in terms of a grid tessellation where the domain is a set of
congruent rectangles. Subtypes of CV_DiscreteSurfaceCoverage that may be used in classification systems
are illustrated in Figure 7. Different grid tessellations may be used. One type of quadrilateral grid is an
equal-cell-size regular grid with a linear traversal sequence order; however, other grid organizations may be
used, such as a variable cell size “QuadTree” grid with a Morton traversal sequence order.
CoverageFunction CV_SurfaceValuePair
CV_DiscreteSurfaceCoverage (from discrete coverages)
1.*
(from discrete coverages)
+element
+collection + geometry : GM_Surface
+ value : Record
CL_ClassifiedSurface
+ tessellationGeometry [0.1] : CL_TessellationGeometry
+ sequenceRule : CV_SequenceType
<>
<> CV_SequenceType
(from quadrilateral grid)
CL_TessellationGeometry
+ linear
+ quadrilateral grid
+ boustrophedonic
+ hexagonal grid
+ CantorDiagonal
+ TIN
+ spiral
+ polygon network
+ Morton
+ Thiessen Polygon network
+ Hilbert
Figure 7 — Discrete surface coverage types
5.5.4 Attributes
A geometry value pair consists of a spatial, temporal or spatiotemporal object together with a record of feature
attribute values. The value set for a classified data discrete coverage is represented as a collection of records
with a common schema. The attribute rangeType of the object CV_Coverage describes the range of the
coverage. The application schema for a particular classification system defines the record structure used to
contain attribute values for that application system.
Attributes may be references to coded values in a catalogue of legend classes. A legend class corresponds to
a feature type that is defined as part of a classification scheme. That is, the feature type is constrained to be
12 © ISO 2009 – All rights reserved
part of the classification system that imposes some overall structure on the set of legend classes. The
structure can be organized in either a priori or a posteriori manner, as described in Annex B. Figure 8 shows
the relationship of a legend class to a feature type. The catalogue of legend classes may be held in a register
as described in Clause 6.
A CL_LegendClass is a metaclass that is a subtype of the general GF_AttributeType metaclass from the
GeneralFeatureModel of ISO 19109. It is implemented by a reference to a catalogue of legend classes
through a CL_FeatureTypeReference. Feature attribute metadata can be associated with a legend class.
MD_Metadata
(from metadata entity set information)
+ fileIdentifier[0.1] : CharacterString
+ language[0.1] : CharacterString
+ characterSet[0.1] : MD_CharacterSetCode = “utf8”
+ parentIdentifier[0.1] : CharacterString
+ hierarchyLevel[0.*] : MD_ScopeCode = “dataset”
<>
+ hierarchyLevelName[0.*] : CharacterString
0.1
+characterize
GF_AttributeType
+ contact[1.*] : CI_ResponsibleParty
(from general feature model)
+ dateStamp : Date
+ metadataStandardName[0.1] : CharacterString + valueType : TypeName
AttributeOfAttribute
+ metadataStandardVersion[0.1] : CharacterString + valueDomain : CharacterString
+ dataSet[0.1] : CharacterString + cardinality : Multiplicity
+characterizeBy
0.* +featureAttributeMetadata
0.*
<>
CL_LegendClass
<>
+featureType
CL_FeatureTypeReference
+ identifier : CharacterString
+ sourceCitation [0.1] : CI_Citation
Figure 8 — Relationship of legend class to feature type
5.6 A classification data set
A classification data set can include a discrete coverage, together with associated metadata. Figure 9
illustrates an overview of a content model for classified discrete coverage data. This is a specialization of a
corresponding illustration from ISO/TS 19129. Metadata elements consist of two types — those that describe
the context of the data and those that describe the content of the data. Both context and content metadata are
defined in ISO 19115. Imagery metadata from ISO 19115-2 can be used in some cases to describe the
method by which classification data has been acquired. The geometric structure and attribute data consists of
spatial referencing and a value set of data elements.
Figure 9 — Overview of
...
INTERNATIONAL ISO
STANDARD 19144-1
First edition
2009-08-15
Geographic information — Classification
systems —
Part 1:
Classification system structure
Information géographique — Systèmes de classification —
Partie 1: Structure de système de classification
Reference number
©
ISO 2009
PDF disclaimer
This PDF file may contain embedded typefaces. In accordance with Adobe's licensing policy, this file may be printed or viewed but
shall not be edited unless the typefaces which are embedded are licensed to and installed on the computer performing the editing. In
downloading this file, parties accept therein the responsibility of not infringing Adobe's licensing policy. The ISO Central Secretariat
accepts no liability in this area.
Adobe is a trademark of Adobe Systems Incorporated.
Details of the software products used to create this PDF file can be found in the General Info relative to the file; the PDF-creation
parameters were optimized for printing. Every care has been taken to ensure that the file is suitable for use by ISO member bodies. In
the unlikely event that a problem relating to it is found, please inform the Central Secretariat at the address given below.
© ISO 2009
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying and microfilm, without permission in writing from either ISO at the address below or
ISO's member body in the country of the requester.
ISO copyright office
Case postale 56 • CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax + 41 22 749 09 47
E-mail copyright@iso.org
Web www.iso.org
Published in Switzerland
ii © ISO 2009 – All rights reserved
Contents Page
Foreword .iv
Introduction.v
1 Scope.1
2 Conformance .1
2.1 Classes .1
2.2 Conformance of a classification system.1
2.3 Conformance of a register of classifiers.1
2.4 Representation of classification results .1
3 Normative references.1
4 Terms, definitions and abbreviated terms.2
4.1 Terms and definitions .2
4.2 Abbreviated terms.4
5 Classification systems.5
5.1 Concept .5
5.2 Classification and legend .7
5.3 Hierarchical versus non-hierarchical systems.8
5.4 A priori and a posteriori classification systems .8
5.5 Structure of classified data .9
5.6 A classification data set.13
6 Management of classifiers .14
6.1 General .14
6.2 Concept dictionary register for a classification scheme .15
6.3 Management of classifiers through registration.15
6.4 Register structure.15
Annex A (normative) Abstract test suite .24
Annex B (informative) A priori and a posteriori classification systems .26
Bibliography.30
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards bodies
(ISO member bodies). The work of preparing International Standards is normally carried out through ISO
technical committees. Each member body interested in a subject for which a technical committee has been
established has the right to be represented on that committee. International organizations, governmental and
non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the
International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization.
International Standards are drafted in accordance with the rules given in the ISO/IEC Directives, Part 2.
The main task of technical committees is to prepare International Standards. Draft International Standards
adopted by the technical committees are circulated to the member bodies for voting. Publication as an
International Standard requires approval by at least 75 % of the member bodies casting a vote.
Attention is drawn to the possibility that some of the elements of this document may be the subject of patent
rights. ISO shall not be held responsible for identifying any or all such patent rights.
ISO 19144-1 was prepared jointly by the Food and Agriculture Organization (FAO) of the United Nations and
Technical Committee ISO/TC 211, Geographic information/Geomatics under a cooperative agreement
between the two organizations.
ISO 19144 consists of the following parts, under the general title Geographic information — Classification
systems:
⎯ Part 1: Classification system structure
The following part is under preparation:
⎯ Part 2: Land cover classification system (LCCS)
iv © ISO 2009 – All rights reserved
Introduction
This part of ISO 19144 is based on publications of the Food and Agriculture Organization (FAO) of the United
[1][2]
Nations . The first in a series of International Standards related to geographic classification systems, it
defines the structure of such systems, together with the mechanism for defining and registering classifiers.
Since there are many different possible application areas, there is no single classification system that will
serve all needs. The method by which classifiers are defined depends upon the application area. In addition,
the classifiers used within a particular application area might not be adequate for all situations encountered
within that application area and could need to be augmented over time. To facilitate extension of the set of
classifiers in a particular application area, classifiers are registered in a register structure compliant with
ISO 19135. This allows the set of classifiers to be maintained. The use of the ISO 19135 registration
mechanism allows for separate registers to be defined for different sets of classifiers within multiple
information communities, thereby satisfying application needs. This approach allows for independence
between information communities, but also allows relationships to be developed between different
classification systems that potentially allow the conversion, or partial conversion, of data from one
classification system to another, or the fusion of data from two separate sources.
The concept of classification systems is well known in the geographic information community. A classification
system can be used to subdivide any geographic area into small units, each of which carries an identifier that
describes its type. The results can then be represented as a discrete coverage as described in ISO 19123.
Many such classification systems can be defined to address any geographic area. Different application areas
and different information communities can define their own classification systems. However, if the
classification system is defined in a compatible way, interaction between different information communities
becomes possible. In addition, in a particular application area, it is desirable that there be a few
well-established classification systems, and that these themselves be standardized within information
communities.
This part of ISO 19144 describes the common structure, while subsequent parts will allow for the
standardization of specific classification systems.
A coverage is a function that returns values from its range for any direct position within its spatial, temporal or
spatiotemporal domain. A discrete coverage is a function that returns the same feature attribute values for
every direct position within any single spatial object, temporal object or spatiotemporal object in its domain.
The domain is an area covered by the coverage function, and the discrete coverage breaks that area down
into a set of spatial, temporal or spatiotemporal objects. The geometry of the discrete coverage used to
represent the results of applying a classification system can be any type of discrete coverage — for example,
a set of polygons fitted together like a jig-saw puzzle, a set of grid cells, or a set of points or curves.
A classification system consists of a set of classifiers. These classifiers may be algorithmically defined, or
established according to a set of classification system definitions. The classifiers are
application-area-dependent and are or will be defined in the other parts of ISO 19144 or other standards or
publications. A register allows for the maintenance of a set of classifiers for a particular application area. A
spatial, temporal or spatiotemporal object defined in terms of a set of classifiers is a classified object.
There is a commonality between conventional geographic features and classified objects. A feature is defined
in ISO 19101 as an abstraction of real world phenomena. An example of a class of feature is a building, and a
particular building, e.g. the UN building in New York, is an instance of a feature class. Conventional
geographic features are atomic units that are assembled to build one type of geographic information data set.
A classification system works in the opposite manner, from the top down, by successively decomposing the
whole within a coverage area. Classified objects are features, in that they are an abstraction of a real world
phenomena, but classified objects are not atomic, because they are necessarily related to each other by the
classifiers that decompose the whole. In a simple example of a classification system, the earth as a whole can
be covered by either “land” or “water”, and two classifiers can be defined partitioning the attribute range into
two, identifying objects as being either land or water. Any particular area on the earth, corresponding to a
classified object, would be of type “land” or “water”.
ISO 19135 specifies that a technical standard be required to define the item classes in any conformant
register. This part of ISO 19144 defines schemas for registers conformant to ISO 19135 and serves as the
technical standard that defines the item classes required for the registration of classifiers. It establishes a set
of rules for specifying definitions that can be used in a particular context to establish classified objects.
Registers of classifiers can serve as sources of reference for similar registers established by other geographic
information communities as part of a system of cross-referencing. Cross-referencing between respective items
in registers of classifiers might be difficult in cases where the structure of registers differs between information
communities. This part of ISO 19144 can serve as a guide for different information communities for the
development of compatible registers that can support a system of classifier cross-referencing.
The structure of a classification system together with the mechanism of defining and registering classifiers
defined in this part of ISO 19144 is general and can be applied to many different information-community-
defined classification systems, including soil, landform, vegetation, urbanization and systems for
understanding biodiversity and climate change. The use of this document will allow the relationship between
different classification systems to be described.
vi © ISO 2009 – All rights reserved
INTERNATIONAL STANDARD ISO 19144-1:2009(E)
Geographic information — Classification systems —
Part 1:
Classification system structure
1 Scope
This part of ISO 19144 establishes the structure of a geographic information classification system, together
with the mechanism for defining and registering the classifiers for such a system. It specifies the use of
discrete coverages to represent the result of applying the classification system to a particular area and defines
the technical structure of a register of classifiers in accordance with ISO 19135.
The structure can be used to develop specific classification systems that address particular application areas,
specified in other parts of ISO 19144.
2 Conformance
2.1 Classes
Three conformance classes are identified in this part of ISO 19144.
2.2 Conformance of a classification system
Any classification system for which conformance to this part of ISO 19144 is claimed shall be in accordance
with Annex A (see A.2).
2.3 Conformance of a register of classifiers
Any register of classifiers for which conformance to this part of ISO 19144 is claimed shall be in accordance
with Annex A (see A.3) and ISO 19135:2005, A.1.
2.4 Representation of classification results
Any legend of classifiers for which conformance to this part of ISO 19144 is claimed shall be in accordance
with Annex A (see A.4).
3 Normative references
The following referenced documents are indispensable for the application of this document. For dated
references, only the edition cited applies. For undated references, the latest edition of the referenced
document (including any amendments) applies.
ISO/TS 19103:2005, Geographic information — Conceptual schema language
ISO 19110:2005, Geographic information — Methodology for feature cataloguing
ISO 19115, Geographic information — Metadata
ISO 19123, Geographic information — Schema for coverage geometry and functions
ISO 19135:2005, Geographic information — Procedures for item registration
4 Terms, definitions and abbreviated terms
For the purposes of this document, the following terms, definitions and abbreviated terms apply.
4.1 Terms and definitions
4.1.1
a posteriori classification
classification (4.1.4) scheme based upon definition of classes after clustering the field samples collected
NOTE 1 Taken from FAO LCCS version 2 (see Reference [2]).
NOTE 2 See Annex B for an examination of this and a priori classification (4.1.2).
4.1.2
a priori classification
classification (4.1.4) scheme structured so that the classes are abstract conceptualizations of the types
actually occurring
NOTE 1 Taken from FAO LCCS version 2 (see Reference [2]).
NOTE 2 The approach is based upon the definition of classes before any data collection actually takes place.
NOTE 3 See Annex B for an examination of this and a posteriori classification (4.1.1).
4.1.3
classified object
spatial object, temporal object or spatiotemporal object assigned to a specific legend class (4.1.16)
4.1.4
classification
abstract representation of real world phenomena using classifiers (4.1.6)
4.1.5
classification system
system for assigning objects to classes
4.1.6
classifier
definition used to assign objects to legend classes (4.1.16)
NOTE Classifiers can be defined algorithmically or according to a set of classification system (4.1.5) specific rules.
4.1.7
coverage
feature (4.1.10) that acts as a function to return values from its range (4.1.17) for any direct position within its
spatial, temporal or spatiotemporal domain (4.1.9)
[ISO 19123:2005]
EXAMPLE Raster image, polygon overlay, digital elevation matrix.
NOTE A coverage is a feature that has multiple values for each attribute type, where each direct position within the
geometric representation of the feature has a single value for each attribute type.
2 © ISO 2009 – All rights reserved
4.1.8
discrete coverage
coverage (4.1.7) that returns the same feature attribute (4.1.11) values for every direct position within any
single spatial object, temporal object, or spatiotemporal object in its domain (4.1.9)
[ISO 19123:2005]
NOTE The domain of a discrete coverage consists of a finite set of spatial, temporal, or spatiotemporal objects.
4.1.9
domain
well-defined set
[ISO/TS 19103:2005]
NOTE Domains are used to define the domain and range (4.1.17) of operators and functions.
4.1.10
feature
abstraction of real world phenomena
[ISO 19101:2002]
EXAMPLE The phenomenon “Eiffel Tower” can be classified with other similar phenomena into a feature type
“tower”.
NOTE A feature can occur as a type or an instance. In this part of ISO 19144, type is meant unless otherwise
specified.
4.1.11
feature attribute
characteristic of a feature (4.1.10)
[ISO 19101:2002]
4.1.12
feature concept dictionary
dictionary that contains definitions of, and related descriptive information about concepts that may be specified
in detail in a feature (4.1.10) catalogue
1)
[ISO 19126:— ]
4.1.13
identifier
linguistically independent sequence of characters capable of uniquely and permanently identifying that with
which it is associated
[ISO 19135:2005]
4.1.14
item class
set of items with common properties
[ISO 19135:2005]
NOTE “Class” is used in this context to refer to a set of instances, not the concept abstracted from that set of
instances.
1) To be published.
4.1.15
legend
application of a classification (4.1.4) in a specific area using a defined mapping scale and specific data set
NOTE Taken from FAO LCCS version 2 (see Reference [2]).
4.1.16
legend class
class resultant from the application of a classification (4.1.4) process
NOTE The result of a classification process is termed legend class in this part of ISO 19144 in order to avoid
confusion with the term “class” as used in UML modelling.
4.1.17
range
〈coverage〉 set of feature attribute (4.1.11) values associated by a function with the elements of the
domain (4.1.9) of a coverage (4.1.7)
[ISO 19123:2005]
4.1.18
register
set of files containing identifiers (4.1.13) assigned to items with descriptions of the associated items
[ISO 19135:2005]
4.1.19
registry
information system on which a register (4.1.18) is maintained
[ISO 19135:2005]
4.1.20
technical standard
standard containing the definitions of item classes (4.1.14) requiring registration
[ISO 19135:2005]
4.1.21
vector geometry
representation of geometry through the use of constructive geometric primitives
[ISO 19107:2003]
4.2 Abbreviated terms
4.2.1 General
CRS Coordinate Reference System
LCCS Land Cover Classification System
UML Unified Modelling Language
4.2.2 Notation
The conceptual schema specified in this part of ISO 19144 is described using the Unified Modelling Language
(UML), following the guidance of ISO/TS 19103.
4 © ISO 2009 – All rights reserved
Several model elements used in this schema are defined in other standards in the ISO 19100 series. By
2)
convention within this suite of International Standards, names of UML classes , with the exception of basic
data type classes, include a two-letter prefix that identifies the International Standard and the UML package in
which the class is defined.
UML classes defined in this part of ISO 19144 have the two-letter prefix “CL”.
Table 1 lists the other International Standards and the packages in which UML classes used in this part of
ISO 19144 have been defined.
Table 1 — Sources of externally defined UML classes
International
Prefix Package
Standard
CV ISO 19123 Coverage core and discrete coverages
DS ISO 19115 Metadata application information
GF ISO 19109 General feature model
GM ISO 19107 Geometry root
MD ISO 19115 Metadata entity set information
MI ISO 19115-2 Metadata entity set imagery
RE ISO 19135 Procedures for registration
SC ISO 19111 Spatial referencing by coordinates
TM ISO 19108 Temporal objects
5 Classification systems
5.1 Concept
A discrete coverage returns the same feature attribute for every direct position within any single geometric
object in its spatiotemporal domain. The spatiotemporal domain consists of a set of geometric objects that
together form the coverage.
EXAMPLE 1 The discrete coverage of postal zones within a country: each zone has a different code and it is not
possible to interpolate between these codes. Nevertheless, there could be a high level relationship between the codes.
The small country of Monaco is divided into five quartiers: “Moneghetti”, “La Condamine”, “Fontvieille”, “Monaco-Ville” and
“Monte-Carlo”. These political jurisdictions completely cover the area of the country. The area of Monaco can be
represented as a discrete coverage with five spatial objects where each object has the geometry of a polygon. The
attribute value for each spatial object is the name of the political jurisdiction. See Figure 1.
2) There is potential confusion between the use of the term “class” as used in UML, “class” as used in a classification
scheme and “item class” as used in the procedures for registration. Class as used in a classification scheme is termed
legend class in this part of ISO 19144.
Moneghetti La Condamine
Monte-Carlo
Monaco
Fontvieille Monaco-Ville
Figure 1 — Example of discrete coverage with polygon geometry
The geometry of the spatial objects associated with a discrete coverage can also be grid cells within a grid
structure. Each of the grid cells may carry an attribute.
EXAMPLE 2 A discrete coverage with grid geometry of the same area as used in the previous example. See Figure 2.
The figure legend identifies the instances of the attribute values that actually exist in the data.
Monte-Carlo
Monaco
Moneghetti
Monaco-Ville
Fontvieille
La Condamine
Figure 2 — Example of discrete coverage with grid geometry
NOTE The discrete coverages illustrated by the above examples are simple because only one relatively simple
attribute has been used. However, in reality the attributes for each of the coverage spatial objects can be very complex. In
order to describe land cover it is necessary to integrate a large number of descriptive parameters related to soil, biology
and density into a comprehensive land cover classification system. Such a classification system is of course
application-area-dependent. An oceanographer will have a different classification system than that of a meteorologist.
Classification systems can vary widely in different application areas, but for similar application areas there needs to be
some commonality in order for data from different sources to be used together.
6 © ISO 2009 – All rights reserved
A classification system allows one to define classifiers in order to partition the attribute range of a discrete
coverage to establish classified objects.
5.2 Classification and legend
Classification is an abstract representation of real world phenomena (i.e. the situation in the field) using
classifiers. A classification is a systematic framework with the names of the classes and the definitions used to
distinguish them, and the relation between classes. Classification thus necessarily involves definition of class
boundaries that must be clear and based upon objective criteria.
A classification system shall be
⎯ scale independent, meaning that the classes at all levels of the system shall be applicable at any scale or
level of detail, and
⎯ source independent, implying that it is independent of the means used to collect information.
NOTE 1 Scale independence and source independence exist in the general case of a classification system. When the
system is applied to real data, i.e. when a legend is created, the scale and source limitations of the data can create
restrictions on the legend.
A legend is the application of a classification in a specific area using a defined mapping scale and specific
data set. Therefore, a legend may contain only a proportion, or subset, of all possible classes of the
classification.
A legend shall be
⎯ scale dependent, and
⎯ source dependent.
EXAMPLE 1 Scale dependence: some elements might be too small to be delineated independently when collected at
a particular scale.
EXAMPLE 2 Source dependence: elements collected from different sources, such as an aerial photograph and a
satellite false colour composite image, could be different.
A legend is a selection of a set of the classifiers from all of the possible classifiers within a classification
system. A legend could correspond to the classifiers that apply to a single data set or it could be applicable to
a number of related data sets.
NOTE 2 A legend is analogous to the set of features within a feature catalogue, which is a subset of all of the possible
features within a feature concept dictionary.
EXAMPLE 3 The derivation of legend classes from a classification system: when applying the reference classification
system to a given geographic area with a specific data type, only a subset of the classes from the reference system will
occur in the real data. This set of classes is the legend with which the mapping system will be carried out. The reference
system in Figure 3 is shown as a matrix with two parameters: the height of plants, ranging from 5 m to greater than 50 m,
and the percentage crown cover, ranging from less than 10 % to 100 %. The classes are numbered with a code number
from 1 to 100. In this specific case, the legend will have only four reference classification classes, numbers 33, 47, 78 and
83.
Reference Classification System
>50 m 30 m >5 m
12 3 4
100%
DEFINED
GEOGRAPHIC
47 AREA
50%
given scale and data type
97 98 99 100
10%
mapping
derived
units
legend
[2]
Figure 3 — Legend as application of classification in a particular area
5.3 Hierarchical versus non-hierarchical systems
There are two basic forms of classification system: hierarchical and non-hierarchical.
a) Hierarchical classification systems
⎯ Classifiers shall be defined so that all classes at a specific level of the hierarchy are mutually
exclusive.
⎯ Criteria used to define a classifier at one level of a hierarchical classification shall not be repeated at
another level (e.g. criteria used to define a classifier at a lower level shall not be duplicated to a
higher level of the hierarchy).
NOTE 1 Most classification systems are hierarchically structured. Such a classification offers greater consistency,
owing to its ability to accommodate different levels of information — starting with structured broad-level classes —
which allow further systematic subdivision into more detailed subclasses. At the higher levels of the classification
system, few diagnostic criteria are used; whereas at lower levels, the number of diagnostic criteria increases.
b) Non-hierarchical classification systems
Classifiers shall be defined so that all classes are mutually exclusive.
NOTE 2 Non-hierarchical classification systems apply to only a single level of information, e.g. a single scale or
type of information.
5.4 A priori and a posteriori classification systems
Two approaches are defined for establishing classification systems: a priori and a posteriori classifications.
The first is based on the establishment of a classification scheme before the collection of the data to which it is
applied, whereas the second is based on the establishment of classes based on the common properties
identified in the data collected.
8 © ISO 2009 – All rights reserved
NOTE Both approaches have applications. An a posteriori approach to classification provides additional flexibility that
allows the classification scheme to more closely match the data. However, it can make it more difficult to compare data
collected and classified separately. If standardized rules are used to establish classes in an a posteriori classification
system, then these rules can be used to generate broader sets of classes that encompass multiple separate
classifications. The a priori and a posteriori classification system approaches are discussed in Annex B.
5.5 Structure of classified data
5.5.1 Systematic arrangement
Classification is the process of systematic arrangement of objects into groups on the basis of their attributes.
The set of classes established in a classification system allows one to organize the attributes of geographic
features so that the attribute value for a feature instance is the name of a class from the classification system.
The groups can be directly defined or parametric rules can be used to generate the groups. A subset of the
set of groups in a specific area becomes a legend.
NOTE The concept of classification is very general and is widely used in many scientific and other disciplines.
Most geographic attributes are distributed spatially. That is, instances of geographic attributes normally have
different values at different spatial locations. Geographic information with discrete attributes can be
represented as a discrete coverage or with vector geometry describing a boundary-defined spatial data set
(vector data). Since the same real-world phenomenon may be represented using either coverage or vector
data, the systematic arrangement of objects into groups forming the classification can be applied to both
coverage and boundary-defined spatial vector data.
5.5.2 Representation using discrete coverages
The result of classifying an area is a legend that may be represented as a type of discrete coverage, as
specified in ISO 19123. The range of a discrete coverage is a set of feature attribute values represented as a
set of records with a common schema defined by the classification system. The domain consists of a finite
collection of classified objects together with their direct positions. The geometric object and its associated
record form a geometry value pair.
The class CV_Coverage (see ISO 19123) represents a feature type in accordance with the general feature
model from ISO 19109:2005, Figure 4. Class CV_Coverage has three attributes: the domainExtent, the
rangeType and the commonPointRule. In discrete coverages, there is a direct one-to-one relationship where
each CV_GeometryValuePair links to corresponding attributes. Attribute domainExtent describes the extent of
the domain coverage. The data type EX_Extent is defined in ISO 19115. Attribute rangeType describes the
structure and composition of the attribute data record. Attribute rangeType makes use of the data type
RecordType, which is defined in ISO/TS 19103. A RecordType is a metaclass that describes the structure of a
set of records. A RecordType consists of a list of attribute name/data type pairs. A simple list is the most
common form of rangeType. Attribute commonPointRule identifies the procedure to be used for evaluating the
CV_Coverage at a position that falls either on a boundary between geometric objects or within the boundaries
of two or more overlapping geometric objects, where the geometric objects are either CV_DomainObjects or
CV_ValueObjects. The data type CV_CommonPointRule is defined ISO 19123.
Associated with a CV_Coverage is a specification of the coordinate reference system (CRS) to which the
objects in the domain are referenced. The coordinate reference system is defined in ISO 19111.
Also associated with the CV_Coverage class are the CV_Domain and the CV_AttributeValues classes. The
CV_Coverage links the set of domain objects to the set of records containing the attribute values. A
CV_DomainObject can be any spatial or temporal or spatiotemporal object. The domain objects of a
CV_DiscreteSurfaceCoverage are constrained to be instances of GM_Surface. There is one instance of
CV_AttributeValues (i.e. one data record) for each instance of CV_DomainObject (i.e. each grid cell or
polygon area).
See Figure 4.
<>
GF_FeatureType
(from general feature model)
<>
<>
CV_Coverage
(from coverage core)
CoordinateReferenceSystem
+ domainExtent[1.*] : EX_Extent
0.*
+ rangeType : RecordType
+ commonPointRule : CV_CommonPointRule
+collection +collection
Range
Domain
1 +CRS 1.* +domainElement 0.* +rangeElement
<>
CV_AttributeValues
CV_DomainObject
SC_CRS (from coverage core) (from coverage core)
(from spatial referencing by coordinates)
+ values : Record
+spatialElement +temporalElement
SpatialComposition TemporalComposition
0.* 0.*
<>
TM_GeometricPrimitive
GM_Object (from temporal objects)
(from geometry root)
Figure 4 — Elements of a coverage
CV_DiscreteCoverage is the subclass that returns the same record of feature attribute values for any direct
position within a single CV_DomainObject in its domain. Each geometry value pair consists of a domain object
(e.g. a grid cell or a polygon) and a record of feature attribute values. See Figure 5. In discrete coverages
there is a direct one-to-one relationship where each CV_GeometryValuePair is composed of a
CV_DomainObject such as a GM_Object and a CV_AttributeValues value.
10 © ISO 2009 – All rights reserved
<>
CV_Coverage
(from coverage core)
+ domainExtent[1.*] : EX_Extent
+ rangeType : RecordType
+ commonPointRule : CV_CommonPointRule
CV_DiscreteCoverage
(from discrete coverages)
0.* +collection
CoverageFunction
1.* +element
CV_GeometryValuePair
(from coverage core)
+ geometry : CV_DomainObject
+ value : Record
Figure 5 — Elements of a discrete coverage
The type of discrete coverage is based on the type of geometric object in the spatial domain. ISO 19123:2005,
Clause 6, identifies five types of discrete coverage with different geometries:
⎯ a discrete point coverage, consisting of a set of independent points;
⎯ a grid point coverage, consisting of a set of grid points;
⎯ a discrete curve coverage, consisting of a set of curves;
⎯ a discrete surface coverage, consisting of a set of surfaces, typically GM_Surface objects (polygons) or
congruent rectangles (grid cells) or regular hexagons, but possibly TIN objects or Thiessen Polygon
objects;
⎯ a discrete solid coverage, consisting of a set of solid volumes.
A classification system can make use of any type of geometric object in its spatial domain. It is possible to
establish classifiers to identify the range values for any type of discrete coverage. Discrete coverages that
partition an area are addressed by discrete surface coverage.
5.5.3 Discrete surface coverage
A classification system that makes use of a coverage whose range partitions an area shall make use of the
CV_Discrete surface coverage in accordance with ISO 19123:2005, 6.8. The use of GM_Surface as the
geometry element is illustrated in Figure 6.
CoverageFunction CV_GeometryValuePair
CV_DiscreteCoverage (from coverage core)
+collection +element
(from discrete coverages)
0.* 1.*
+ geometry : CV_DomainObject
+ value : Record
CoverageFunction
CV_SurfaceValuePair
CV_DiscreteSurfaceCoverage
+collection +element
(from discrete coverages)
(from discrete coverages)
1.*
+ geometry : GM_Surface
Figure 6 — Discrete surface coverage
The surfaces that constitute the domain of a discrete surface coverage can be mutually exclusive and
exhaustively partition an area or can be independent sub-areas. The type of coverage geometry used often
depends upon the method by which data was collected. Natural phenomena tend to have irregular boundaries
and are often represented as a set of congruent polygons. Some phenomena, types of sensor data or the
results of calculations tend to be organized in terms of a grid tessellation where the domain is a set of
congruent rectangles. Subtypes of CV_DiscreteSurfaceCoverage that may be used in classification systems
are illustrated in Figure 7. Different grid tessellations may be used. One type of quadrilateral grid is an
equal-cell-size regular grid with a linear traversal sequence order; however, other grid organizations may be
used, such as a variable cell size “QuadTree” grid with a Morton traversal sequence order.
CoverageFunction CV_SurfaceValuePair
CV_DiscreteSurfaceCoverage (from discrete coverages)
1.*
(from discrete coverages)
+element
+collection + geometry : GM_Surface
+ value : Record
CL_ClassifiedSurface
+ tessellationGeometry [0.1] : CL_TessellationGeometry
+ sequenceRule : CV_SequenceType
<>
<> CV_SequenceType
(from quadrilateral grid)
CL_TessellationGeometry
+ linear
+ quadrilateral grid
+ boustrophedonic
+ hexagonal grid
+ CantorDiagonal
+ TIN
+ spiral
+ polygon network
+ Morton
+ Thiessen Polygon network
+ Hilbert
Figure 7 — Discrete surface coverage types
5.5.4 Attributes
A geometry value pair consists of a spatial, temporal or spatiotemporal object together with a record of feature
attribute values. The value set for a classified data discrete coverage is represented as a collection of records
with a common schema. The attribute rangeType of the object CV_Coverage describes the range of the
coverage. The application schema for a particular classification system defines the record structure used to
contain attribute values for that application system.
Attributes may be references to coded values in a catalogue of legend classes. A legend class corresponds to
a feature type that is defined as part of a classification scheme. That is, the feature type is constrained to be
12 © ISO 2009 – All rights reserved
part of the classification system that imposes some overall structure on the set of legend classes. The
structure can be organized in either a priori or a posteriori manner, as described in Annex B. Figure 8 shows
the relationship of a legend class to a feature type. The catalogue of legend classes may be held in a register
as described in Clause 6.
A CL_LegendClass is a metaclass that is a subtype of the general GF_AttributeType metaclass from the
GeneralFeatureModel of ISO 19109. It is implemented by a reference to a catalogue of legend classes
through a CL_FeatureTypeReference. Feature attribute metadata can be associated with a legend class.
MD_Metadata
(from metadata entity set information)
+ fileIdentifier[0.1] : CharacterString
+ language[0.1] : CharacterString
+ characterSet[0.1] : MD_CharacterSetCode = “utf8”
+ parentIdentifier[0.1] : CharacterString
+ hierarchyLevel[0.*] : MD_ScopeCode = “dataset”
<>
+ hierarchyLevelName[0.*] : CharacterString
0.1
+characterize
GF_AttributeType
+ contact[1.*] : CI_ResponsibleParty
(from general feature model)
+ dateStamp : Date
+ metadataStandardName[0.1] : CharacterString + valueType : TypeName
AttributeOfAttribute
+ metadataStandardVersion[0.1] : CharacterString + valueDomain : CharacterString
+ dataSet[0.1] : CharacterString + cardinality : Multiplicity
+characterizeBy
0.* +featureAttributeMetadata
0.*
<>
CL_LegendClass
<>
+featureType
CL_FeatureTypeReference
+ identifier : CharacterString
+ sourceCitation [0.1] : CI_Citation
Figure 8 — Relationship of legend class to feature type
5.6 A classification data set
A classification data set can include a discrete coverage, together with associated metadata. Figure 9
illustrates an overview of a content model for classified discrete coverage data. This is a specialization of a
corresponding illustration from ISO/TS 19129. Metadata elements consist of two types — those that describe
the context of the data and those that describe the content of the data. Both context and content metadata are
defined in ISO 19115. Imagery metadata from ISO 19115-2 can be used in some cases to describe the
method by which classification data has been acquired. The geometric structure and attribute data consists of
spatial referencing and a value set of data elements.
Figure 9 — Overview of a content model for classified discrete coverage data
The model shown in Figure 10 illustrates the relationship between a classification collection and associated
metadata comprising a data set. The class CL_ClassificationCollection is a collection class that links the
coverage to the metadata.
DS_DataSet
(from metadata application information)
+describes +has
MD_Metadata
CL_DataSet
0.* 1.* (from metadata entity set information)
Associated metadata
CL_ClassificationCollection
+collection 0.*
MI_Metadata
(from metadata entity set imagery)
CoverageFunction
+element 1.*
CV_SurfaceValuePair
(from discrete coverages)
+ geometry : GM_Surface
+ value : Record
Figure 10 — Classification data set content elements
6 Management of classifiers
6.1 General
A classified object is a subtype of a feature attribute that is constrained to work within a particular classification
scheme. The process of establishing a classification system subdivides the attribute range of a discrete
coverage. Like a feature attribute, a classification can be predefined with the definition stored within a
dictionary. The dictionary might contain all of the classifiers available within the classification scheme. New
classifiers might be a
...
NORME ISO
INTERNATIONALE 19144-1
Première édition
2009-08-15
Information géographique — Systèmes
de classification —
Partie 1:
Structure de système de classification
Geographic information — Classification systems —
Part 1: Classification system structure
Numéro de référence
©
ISO 2009
DOCUMENT PROTÉGÉ PAR COPYRIGHT
Droits de reproduction réservés. Sauf prescription différente, aucune partie de cette publication ne peut être reproduite ni utilisée sous
quelque forme que ce soit et par aucun procédé, électronique ou mécanique, y compris la photocopie et les microfilms, sans l’accord écrit
de l’ISO à l’adresse ci-après ou du comité membre de l’ISO dans le pays du demandeur.
ISO copyright office
Case postale 56 • CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax + 41 22 749 09 47
E-mail copyright@iso.org
Web www.iso.org
Version française parue en 2012
Publié en Suisse
ii © ISO 2009 – Tous droits réservés
Sommaire Page
Avant-propos .iv
Introduction . v
1 Domaine d’application . 1
2 Conformité . 1
2.1 Classes . 1
2.2 Conformité d’un système de classification . 1
2.3 Conformité d’un registre de classificateurs . 1
2.4 Représentation des résultats de classification . 1
3 Références normatives . 1
4 Termes, définitions et termes abrégés . 2
4.1 Termes et définitions . 2
4.2 Termes abrégés . 4
5 Systèmes de classification . 5
5.1 Concept . 5
5.2 Classification et légende . 7
5.3 Systèmes hiérarchiques contre systèmes non hiérarchiques . 8
5.4 Systèmes de classification a priori et a posteriori . 8
5.5 Structure de données classifiées . 9
5.6 Ensemble de données de classification .13
6 Gestion des classificateurs .14
6.1 Généralités .14
6.2 Registre de dictionnaire de concepts pour un principe de classification .15
6.3 Gestion des classificateurs par l’enregistrement .15
6.4 Structure de registres .16
Annexe A (normative) Suite d’essais abstraits .25
Annexe B (informative) Systèmes de classification a priori et a posteriori.27
Bibliographie .31
Avant-propos
L’ISO (Organisation internationale de normalisation) est une fédération mondiale d’organismes nationaux de
normalisation (comités membres de l’ISO). L’élaboration des Normes internationales est en général confiée aux
comités techniques de l’ISO. Chaque comité membre intéressé par une étude a le droit de faire partie du comité
technique créé à cet effet. Les organisations internationales, gouvernementales et non gouvernementales,
en liaison avec l’ISO participent également aux travaux. L’ISO collabore étroitement avec la Commission
électrotechnique internationale (CEI) en ce qui concerne la normalisation électrotechnique.
Les Normes internationales sont rédigées conformément aux règles données dans les Directives ISO/CEI, Partie 2.
La tâche principale des comités techniques est d’élaborer les Normes internationales. Les projets de Normes
internationales adoptés par les comités techniques sont soumis aux comités membres pour vote. Leur publication
comme Normes internationales requiert l’approbation de 75 % au moins des comités membres votants.
L’attention est appelée sur le fait que certains des éléments du présent document peuvent faire l’objet de droits
de propriété intellectuelle ou de droits analogues. L’ISO ne saurait être tenue pour responsable de ne pas avoir
identifié de tels droits de propriété et averti de leur existence.
L’ISO 19144-1 a été élaborée conjointement par l’Organisation des Nations Unies pour l’Alimentation et
l’Agriculture (FAO) et le comité technique ISO/TC 211, Information géographique/Géomatique, dans le cadre
d’un accord de coopération entre les deux organisations.
L’ISO 19144 comprend les parties suivantes, présentées sous le titre général Information géographique —
Systèmes de classification:
— Partie 1: Structure de système de classification
La partie suivante est en cours d’élaboration:
— Partie 2: Métalangage de couverture du sol (LCML)
iv © ISO 2009 – Tous droits réservés
Introduction
La présente partie de l’ISO 19144 est basée sur des publications de l’Organisation des Nations Unies pour
[1][2]
l’Alimentation et l’Agriculture (FAO) . Première d’une série de Normes internationales se rapportant aux
systèmes de classification géographique, elle définit la structure de ces systèmes, ainsi que le mécanisme de
définition et d’enregistrement de classificateurs.
Comme il existe de nombreux et divers domaines d’application possibles, il n’existe pas de système de
classification unique répondant à tous les besoins. La méthode par laquelle les classificateurs sont définis
dépend du domaine d’application. De plus, les classificateurs utilisés dans un domaine d’application particulier
peuvent ne pas être adaptés à toutes les situations rencontrées dans ce domaine d’application et pourraient
avoir à être augmentés dans le temps. Afin de faciliter l’extension de l’ensemble de classificateurs dans
un domaine d’application particulier, les classificateurs sont enregistrés dans une structure de registres
conforme à l’ISO 19135. Cela permet d’entretenir l’ensemble de classificateurs. L’utilisation du mécanisme
d’enregistrement selon l’ISO 19135 permet de définir des registres séparés pour différents ensembles de
classificateurs dans plusieurs communautés d’informations, en satisfaisant ainsi aux besoins d’application.
Cette approche permet une indépendance entre communautés d’informations, mais permet également de
développer des relations entre différents systèmes de classification permettant potentiellement la conversion,
ou la conversion partielle, de données d’un système de classification à un autre, ou la fusion de données
issues de deux sources distinctes.
Le concept de systèmes de classification est bien connu dans la communauté de l’information géographique.
Un système de classification peut être utilisé pour subdiviser une zone géographique en petites unités, chacune
portant un identifiant qui décrit son type. Les résultats peuvent alors être représentés sous forme d’une couverture
discrète comme décrit dans l’ISO 19123. Nombre de ces systèmes de classification peuvent être définis pour
traiter toute zone géographique. Différents domaines d’application et différentes communautés d’informations
peuvent définir leurs propres systèmes de classification. Cependant, si le système de classification est défini
d’une manière compatible, l’interaction entre différentes communautés d’informations devient possible. En
outre, dans un domaine d’application particulier, il est souhaitable d’avoir peu de systèmes de classification
bien établis et que ceux-ci soient normalisés au sein de communautés d’informations.
La présente partie de l’ISO 19144 décrit la structure commune, alors que les parties suivantes permettront la
normalisation de systèmes de classification spécifiques.
Une couverture est une fonction qui assigne des valeurs à partir de sa plage à toute position directe dans
son domaine spatial, temporel ou spatio-temporel. Une couverture discrète est une fonction qui assigne les
mêmes valeurs d’attributs d’entités à chaque position directe à l’intérieur d’un objet spatial, temporel ou bien
spatio-temporel unique dans son domaine. Le domaine est une zone couverte par la fonction de couverture
et la couverture discrète divise la zone en un ensemble d’objets spatiaux, temporels ou spatio-temporels.
La géométrie de la couverture discrète utilisée pour représenter les résultats de l’application d’un système
de classification peut être de tout type de couverture discrète — par exemple un ensemble de polygones
assemblés comme un puzzle, un ensemble de cellules de grille ou un ensemble de points ou de courbes.
Un système de classification est constitué d’un ensemble de classificateurs. Ces classificateurs peuvent être
définis de façon algorithmique ou être établis selon un ensemble de définitions du système de classification.
Les classificateurs dépendent de la zone d’application et sont ou seront définis dans les autres parties
de l’ISO 19144 ou dans d’autres normes ou publications. Un registre permet d’entretenir un ensemble de
classificateurs pour un domaine d’application particulier. Un objet spatial, temporel ou spatio-temporel défini
en termes d’ensemble de classificateurs est un objet classifié.
Les entités géographiques conventionnelles et les objets classifiés ont un point commun. Une entité est définie
dans l’ISO 19101 comme une abstraction de phénomènes du monde réel. Un bâtiment est un exemple de
classe d’entité et un bâtiment particulier, par exemple le bâtiment de l’ONU à New York, est une instance
de classe d’entité. Les entités géographiques conventionnelles sont des unités atomiques assemblées pour
construire un type d’ensemble de données d’informations géographiques.
Un système de classification fonctionne de la manière opposée, de manière descendante, en décomposant
successivement le tout en une zone de couverture. Les objets classifiés sont des entités, en ce qu’elles sont
une abstraction de phénomènes du monde réel, mais les objets classifiés ne sont pas atomiques, car ils sont
nécessairement liés les uns aux autres par les classificateurs qui décomposent le tout. Dans un exemple
simple de système de classification, la Terre en tant que tout peut être recouverte par de la «terre» ou de
«l’eau» et deux classificateurs peuvent être définis en fractionnant la plage d’attributs en deux, en identifiant
les objets comme étant de la terre ou de l’eau. Toute zone particulière de la Terre, correspondant à un objet
classifié, serait de type «terre» ou «eau».
L’ISO 19135 spécifie qu’une norme technique est nécessaire pour définir les classes d’éléments dans tout
registre conforme. La présente partie de l’ISO 19144 définit des schémas pour des registres conformes à
l’ISO 19135 et sert de norme technique pour définir les classes d’éléments nécessaires pour l’enregistrement
de classificateurs. Elle établit un ensemble de règles pour spécifier des définitions pouvant être utilisées dans
un contexte particulier afin d’établir des objets classifiés.
Les registres de classificateurs peuvent servir de sources de référence pour des registres similaires établis par
d’autres communautés d’informations géographiques dans le cadre d’un système de référencement croisé. Le
référencement croisé entre des éléments respectifs dans des registres de classificateurs peut être difficile dans
les cas où la structure des registres diffère d’une communauté d’informations à l’autre. La présente partie de
l’ISO 19144 peut servir de guide pour différentes communautés d’informations concernant le développement
de registres compatibles pouvant prendre en charge un système de référencement croisé de classificateurs.
La structure d’un système de classification ainsi que le mécanisme de définition et d’enregistrement de
classificateurs définis dans la présente partie de l’ISO 19144 sont généraux et peuvent être appliqués à de
nombreux systèmes de classifications, définis par différentes communautés d’information, comprenant le sol,
le relief, la végétation, l’urbanisation et les systèmes visant à analyser la biodiversité et les changements
climatiques. L’utilisation du présent document permettra de décrire la relation entre différents systèmes de
classification.
vi © ISO 2009 – Tous droits réservés
NORME INTERNATIONALE ISO 19144-1:2009(F)
Information géographique — Systèmes de classification —
Partie 1:
Structure de système de classification
1 Domaine d’application
La présente partie de l’ISO 19144 établit la structure d’un système de classification d’informations géographiques,
ainsi que le mécanisme de définition et d’enregistrement des classificateurs pour un tel système. Elle spécifie
l’utilisation de couvertures discrètes pour représenter le résultat de l’application du système de classification à une
zone particulière et définit la structure technique d’un registre de classificateurs, conformément à l’ISO 19135.
La structure peut être utilisée pour développer des systèmes de classification spécifiques traitant de domaines
d’application particuliers, spécifiés dans d’autres parties de l’ISO 19144.
2 Conformité
2.1 Classes
Trois classes de conformité sont identifiées dans la présente partie de l’ISO 19144.
2.2 Conformité d’un système de classification
Tout système de classification, pour lequel la conformité à la présente partie de l’ISO 19144 est revendiquée,
doit être conforme à l’Annexe A (voir en A.2.).
2.3 Conformité d’un registre de classificateurs
Tout registre de classificateurs, pour lequel la conformité à la présente partie de l’ISO 19144 est revendiquée,
doit être conforme à l’Annexe A (voir en A.3.) et à l’ISO 19135:2005, A.1.
2.4 Représentation des résultats de classification
Toute légende de classificateurs, pour laquelle la conformité à la présente partie de l’ISO 19144 est revendiquée,
doit être conforme à l’Annexe A (voir en A.4).
3 Références normatives
Les documents de référence suivants sont indispensables à l’application du présent document. Pour les
références datées, seule l’édition citée s’applique. Pour les références non datées, la dernière édition du
document de référence s’applique (y compris les éventuels amendements).
ISO/TS 19103:2005, Information géographique — Langage de schéma conceptuel
ISO 19110:2005, Information géographique — Méthodologie de catalogage des entités
ISO 19115, Information géographique — Métadonnées
ISO 19123, Information géographique — Schéma de la géométrie et des fonctions de couverture
ISO 19135:2005, Information géographique — Procédures pour l’enregistrement d’éléments
4 Termes, définitions et termes abrégés
Pour les besoins du présent document, les termes, définitions et termes abrégés suivants s’appliquent.
4.1 Termes et définitions
4.1.1
classification a posteriori
principe de classification (4.1.4) basé sur la définition de classes après un regroupement des échantillons
recueillis sur le terrain
NOTE 1 Extrait du Système de classification de l’occupation du sol (LCCS, Land Cover Classification System),
version 2, de la FAO (voir la Référence [2]).
NOTE 2 Voir l’Annexe B pour un examen de cette classification et de la classification a priori (4.1.2).
4.1.2
classification a priori
principe de classification (4.1.4) structuré de façon que les classes soient des conceptualisations abstraites
des types se produisant effectivement
NOTE 1 Extrait du Système de classification de l’occupation du sol (LCCS, Land Cover Classification System),
version 2, de la FAO (voir la Référence [2]).
NOTE 2 Cette approche est basée sur la définition de classes avant que toute collecte de données ne se déroule.
NOTE 3 Voir l’Annexe B pour un examen de cette classification et de la classification a posteriori (4.1.1).
4.1.3
objet classifié
objet spatial, temporel ou spatio-temporel affecté à une classe de légende (4.1.16) spécifique
4.1.4
classification
représentation abstraite à l’aide de classificateurs (4.1.6) de phénomènes survenant dans le monde réel
4.1.5
système de classification
système destiné à assigner des objets à des classes
4.1.6
classificateur
définition utilisée afin d’assigner des objets à des classes de légendes (4.1.16)
NOTE Les classificateurs peuvent être définis de façon algorithmique ou conformément à un ensemble de règles
propres à un système de classification (4.1.5).
4.1.7
couverture
entité (4.1.10) qui agit comme une fonction en assignant des valeurs à partir de sa plage (4.1.17) à toute
position directe dans son domaine (4.1.9) spatial, temporel ou spatio-temporel
[ISO 19123:2005]
EXEMPLE Image matricielle, couverture de polygones, modèle numérique d’altitude.
NOTE Une couverture est une entité qui prend des valeurs multiples selon le type d’attribut, où chaque position
directe dans la représentation géométrique de la caractéristique a une valeur unique pour chaque type d’attribut.
2 © ISO 2009 – Tous droits réservés
4.1.8
couverture discrète
couverture (4.1.7) qui assigne les mêmes valeurs d’attribut d’une entité (4.1.11) à chaque position directe à
l’intérieur d’un objet spatial, temporel ou bien spatio-temporel unique dans son domaine (4.1.9)
[ISO 19123:2005]
NOTE Le domaine d’une couverture discrète se compose d’un ensemble fini d’objets spatiaux, temporels ou
spatio-temporels.
4.1.9
domaine
ensemble bien défini
[ISO/TS 19103:2005]
NOTE Les domaines permettent de définir le domaine et la plage (4.1.17) des opérateurs et des fonctions.
4.1.10
entité
abstraction de phénomènes du monde réel
[ISO 19101:2002]
EXEMPLE Le phénomène «Tour Eiffel» peut être classifié avec d’autres phénomènes similaires dans un type
d’entité «tour».
NOTE Une entité peut se présenter comme un type ou comme une instance. Dans la présente partie de l’ISO 19144,
type est désigné, sauf spécification contraire.
4.1.11
attribut d’une entité
propriété d’une entité (4.1.10)
[ISO 19101:2002]
4.1.12
dictionnaire de concepts d’entités
dictionnaire contenant des définitions de concepts ainsi que les informations de description associées les
concernant, pouvant être spécifiés en détail dans un catalogue d’entités (4.1.10)
1)
[ISO 19126:— ]
4.1.13
identifiant
séquence de caractères linguistiquement indépendante capable d’identifier de manière exclusive et continue
ce à quoi elle est associée
[ISO 19135:2005]
4.1.14
classe d’éléments
ensemble d’éléments ayant des propriétés communes
NOTE 1 Adapté de l’ISO 19135:2005.
NOTE 2 La «classe» est utilisée dans ce contexte afin de faire référence à un ensemble d’instances et non au concept
issu de cet ensemble d’instances.
1) À paraître.
4.1.15
légende
application d’une classification (4.1.4) au sein d’une zone spécifique à l’aide d’une échelle de mise en
correspondance définie et d’un ensemble de données spécifique
NOTE Extrait de la Classification de l’occupation du sol (LCCS, Land Cover Classification System), version 2, de la
FAO (voir la Référence [2]).
4.1.16
classe de légende
classe résultante de l’application d’un processus de classification (4.1.4)
NOTE Le résultat d’un processus de classification est appelé classe de légende dans la présente partie de l’ISO 19144
afin d’éviter toute confusion avec le terme «classe» utilisé dans la modélisation UML.
4.1.17
plage
〈couverture〉 ensemble de valeurs d’attributs d’une entité (4.1.11) associé par une fonction aux éléments du
domaine (4.1.9) d’une couverture (4.1.7)
[ISO 19123:2005]
4.1.18
registre
ensemble de fichiers contenant des identifiants (4.1.13) attribués à des éléments et des descriptions qui leur
sont associées
NOTE Adapté de l’ISO 19135:2005.
4.1.19
bureau d’enregistrement
système d’information dans lequel un registre (4.1.18) est mis à jour
[ISO 19135:2005]
4.1.20
norme technique
norme contenant les définitions des classes d’éléments (4.1.14) requérant un enregistrement
NOTE Adapté de l’ISO 19135:2005.
4.1.21
géométrie vectorielle
représentation d’une géométrie par l’utilisation de primitives géométriques constructives
[ISO 19107:2003]
4.2 Termes abrégés
4.2.1 Généralités
CRS Système de coordonnées de référence (Coordinate Reference System)
LCCS Système de classification d’occupation du sol (Land Cover Classification System)
UML Langage de modélisation unifié (Unified Modelling Language)
4.2.2 Notation
Le schéma conceptuel spécifié dans la présente partie de l’ISO 19144 est décrit à l’aide du Langage de
modélisation unifié (UML, Unified Modelling Language), en suivant les lignes directrices de l’ISO/TS 19103.
4 © ISO 2009 – Tous droits réservés
Plusieurs éléments de modèles utilisés dans ce schéma sont définis dans d’autres normes de la série ISO 19100.
2)
Par convention, dans cette suite de Normes internationales, les noms des classes UML , à l’exception des
classes relatives aux types de données de base, comportent un préfixe à deux lettres qui identifie la Norme
internationale et le «paquetage» UML dans lequel la classe est définie.
Les classes UML définies dans la présente partie de l’ISO 19144 ont le préfixe à deux lettres «CL».
Le Tableau 1 répertorie les autres Normes internationales et les «paquetages» dans lesquels les classes UML
utilisées dans la présente partie de l’ISO 19144 ont été définies.
Tableau 1 — Sources des classes UML définies en dehors du présent document
Préfixe Norme internationale «Paquetage»
CV ISO 19123 Coverage core et couvertures discrètes
DS ISO 19115 Informations d’application de métadonnées
GF ISO 19109 Modèle sémantique
GM ISO 19107 Géométrie
MD ISO 19115 Informations d’ensemble d’entités de métadonnées
MI ISO 19115-2 Images d’ensemble d’entités de métadonnées
RE ISO 19135 Procédures d’enregistrement
SC ISO 19111 Système de références spatiales par coordonnées
TM ISO 19108 Objets temporels
5 Systèmes de classification
5.1 Concept
Une couverture discrète assigne le même attribut d’entité à chaque position directe à l’intérieur d’un objet
géométrique unique dans son domaine spatio-temporel. Le domaine spatio-temporel comprend un ensemble
d’objets géométriques qui forment la couverture.
EXEMPLE 1 La couverture discrète de zones postales dans un pays: chaque zone a un code différent et il n’est pas
possible d’interpoler ces codes. Cependant, il pourrait exister une relation de haut niveau entre les codes. Le petit pays
de Monaco est divisé en cinq quartiers: «Moneghetti», «La Condamine», «Fontvieille», «Monaco-Ville» et «Monte-Carlo».
Ces juridictions politiques couvrent complètement la surface du pays. La superficie de Monaco peut être représentée
par une couverture discrète ayant cinq objets spatiaux, où chaque objet présente la géométrie de polygone. La valeur
d’attribut de chaque objet spatial est le nom de la juridiction politique. Voir Figure 1.
2) Il existe une confusion potentielle entre l’usage du terme «classe» utilisé dans le langage UML, le terme «classe»
utilisé dans un principe de classification, et le terme «classe d’élément» utilisé dans les procédures d’enregistrement. Le
terme classe, qui est utilisé dans un principe de classification, est appelé classe de légende dans la présente partie de
l’ISO 19144.
Moneghetti La Condamine
Monte-Carlo
Monaco
Fontvieille Monaco-Ville
Figure 1 — Exemple de couverture discrète avec une géométrie de polygones
La géométrie des objets spatiaux associés à une couverture discrète peut également être représentée par des
cellules de grille dans une structure de grille. Chacune des cellules de grille peut porter un attribut.
EXEMPLE 2 Une couverture discrète ayant une géométrie de grille de la même superficie que celle utilisée dans
l’exemple précédent. Voir Figure 2. La légende de la figure identifie les instances des valeurs d’attributs qui existent
effectivement dans les données.
Monte-Carlo
Monaco
Moneghetti
Monaco-Ville
Fontvieille
La Condamine
Figure 2 — Exemple de couverture discrète avec une géométrie de grille
NOTE Les couvertures discrètes illustrées par les exemples ci-dessus sont simples car un seul attribut relativement
simple a été utilisé. Cependant, dans la réalité, les attributs pour chacun des objets spatiaux d’une couverture peuvent être
très complexes. Afin de décrire l’occupation du sol, il est nécessaire d’intégrer un grand nombre de paramètres descriptifs
associés au sol, à la biologie et à la densité dans un système de classification d’occupation du sol complet. Un tel système
de classification dépend bien sûr de la zone d’application. Un océanographe aura un système de classification différent
de celui d’un météorologue. Les systèmes de classification peuvent grandement varier dans des zones d’application
différentes mais, pour des zones d’application similaires, quelques points communs doivent exister afin de pouvoir utiliser
ensemble des données issues de sources différentes.
6 © ISO 2009 – Tous droits réservés
Un système de classification permet de définir des classificateurs afin de fractionner la plage d’attributs d’une
couverture discrète pour établir des objets classifiés.
5.2 Classification et légende
La classification est une représentation abstraite de phénomènes du monde réel (c’est-à-dire la situation sur
le terrain) utilisant des classificateurs. Une classification est un cadre systématique accompagné des noms
des classes et des définitions utilisées pour les distinguer et de la relation entre des classes. Une classification
implique donc nécessairement la définition de frontières entre les classes qui doivent être claires et basées sur
des critères objectifs.
Un système de classification doit être
— indépendant de l’échelle, ce qui signifie que les classes à tous les niveaux du système doivent être
applicables à n’importe quelle échelle ou niveau de détail et
— indépendant de la source, ce qui implique qu’il est indépendant des moyens utilisés pour recueillir des
informations.
NOTE 1 L’indépendance vis-à-vis de l’échelle et l’indépendance de la source existent dans le cas général d’un système
de classification. Lorsque le système est appliqué à des données réelles, c’est-à-dire lorsqu’une légende est créée, les
limites des données en termes d’échelle et de source peuvent créer des restrictions pour la légende.
Une légende est l’application d’une classification dans une zone spécifique en utilisant une échelle de mise en
correspondance définie et un ensemble de données spécifique. Ainsi, une légende peut ne contenir qu’une
proportion, ou sous-ensemble, de toutes les classes possibles de la classification.
Une légende doit être
— dépendante de l’échelle et
— dépendante de la source.
EXEMPLE 1 Dépendance vis-à-vis de l’échelle: certains éléments peuvent être trop petits pour être délimités
indépendamment lorsqu’ils sont recueillis à une échelle particulière.
EXEMPLE 2 Dépendance vis-à-vis de la source: les éléments recueillis depuis des sources différentes, comme une
photographie aérienne et une image satellite composite en fausse couleur, peuvent être différents.
Une légende est une sélection d’un ensemble de classificateurs parmi tous les classificateurs possibles dans
un système de classification. Une légende peut correspondre aux classificateurs qui s’appliquent à un seul
ensemble de données ou pourrait s’appliquer à un certain nombre d’ensembles de données associés.
NOTE 2 Une légende est analogue à l’ensemble d’entités dans un catalogue d’entités, qui est un sous-ensemble de
toutes les entités possibles dans un dictionnaire de concepts d’entités.
EXEMPLE 3 Origine des classes de légendes à partir d’un système de classification: lors de l’application d’un système
de classification de référence à une zone géographique donnée avec un type de données spécifique, seul un sous-ensemble
des classes issues du système de référence se produira dans les données réelles. Cet ensemble de classes est la légende
avec laquelle le système de mise en correspondance sera réalisé. Le système de référence de la Figure 3 est représenté
comme une matrice à deux paramètres: la hauteur des plantes, allant de 5 m à plus de 50 m et le pourcentage de couverture
arborée, allant de moins de 10 % à 100 %. Les classes sont numérotées avec un numéro de code allant de 1 à 100. Dans
ce cas spécifique, la légende n’aura que quatre classes de classification de référence, les numéros 33, 47, 78 et 83.
Système de classification de référence
>50 m30 m>5 m
100% 12 4
ZONE
GÉOGRAPHIQUE
47 DÉFINIE
50%
Échelle donnée et type de données
97 98 99 100
10%
Unités de
Légende
mise en
obtenue
correspondance
[2]
Figure 3 — Légende en application d’une classification dans une zone particulière
5.3 Systèmes hiérarchiques contre systèmes non hiérarchiques
Il existe deux formes basiques de systèmes de classification: hiérarchiques et non hiérarchiques.
a) Systèmes de classification hiérarchiques
— Les classificateurs doivent être définis de façon que toutes les classes à un niveau spécifique de la
hiérarchie soient mutuellement exclusives.
— Les critères utilisés pour définir un classificateur à un niveau d’une classification hiérarchique ne doivent
pas être répétés à un autre niveau (par exemple, les critères utilisés pour définir un classificateur à un
niveau inférieur ne doivent pas être dupliqués à un niveau supérieur de la hiérarchie).
NOTE 1 La plupart des systèmes de classification sont structurés hiérarchiquement. Une telle classification offre
une meilleure cohérence en raison de sa capacité à s’adapter à différents niveaux d’informations – à commencer par
les classes structurées dans une large mesure – ce qui permet une subdivision systématique supplémentaire en des
sous-classes plus détaillées. Aux niveaux les plus élevés du système de classification, peu de critères de diagnostic
sont utilisés, alors qu’à des niveaux plus bas, le nombre de critères de diagnostic augmente.
b) Systèmes de classification non hiérarchiques
Les classificateurs doivent être définis de sorte que toutes les classes soient mutuellement exclusives.
NOTE 2 Les systèmes de classification non hiérarchiques ne s’appliquent qu’à un seul niveau d’informations, par
exemple une seule échelle ou un seul type d’informations.
5.4 Systèmes de classification a priori et a posteriori
Deux approches sont définies pour l’établissement de systèmes de classification: les classifications a priori
et a posteriori. La première est basée sur l’établissement d’un principe de classification avant la collecte des
8 © ISO 2009 – Tous droits réservés
données auxquelles il est appliqué, alors que la seconde est basée sur l’établissement de classes fondées sur
les propriétés communes identifiées dans les données recueillies.
NOTE Les deux approches ont des applications. Une approche a posteriori de la classification offre plus de souplesse,
ce qui permet au principe de classification de faire correspondre plus étroitement les données. Cependant, il peut rendre
plus difficile la comparaison de données recueillies et classifiées séparément. Si des règles normalisées sont utilisées
pour établir des classes dans un système de classification a posteriori, alors ces règles peuvent être utilisées pour générer
de plus larges ensembles de classes qui englobent plusieurs classifications séparées. Les approches des systèmes de
classification a priori et a posteriori sont traitées à l’Annexe B.
5.5 Structure de données classifiées
5.5.1 Agencement systématique
La classification est le processus d’agencement systématique d’objets dans des groupes se basant sur leurs
attributs. L’ensemble de classes établies dans un système de classification permet d’organiser les attributs
d’entités géographiques de sorte que la valeur d’attribut pour une instance d’entité soit le nom d’une classe
issue du système de classification. Les groupes peuvent être directement définis ou des règles paramétriques
peuvent être utilisées pour générer les groupes. Un sous-ensemble de l’ensemble de groupes dans une zone
spécifique devient une légende.
NOTE Le concept de classification est très général et est très largement utilisé dans de nombreuses disciplines
scientifiques et autres.
La plupart des attributs géographiques sont répartis dans l’espace. C’est-à-dire que les instances d’attributs
géographiques ont des valeurs différentes à des positions différentes dans l’espace. L’information géographique
avec des attributs discrets peut être représentée comme une couverture discrète ou avec une géométrie
vectorielle décrivant un ensemble de données spatiales définies par des frontières (données vectorielles).
Comme le même phénomène survenant dans le monde réel peut être représenté en utilisant une couverture
ou des données vectorielles, l’agencement systématique d’objets dans des groupes formant la classification
peut être appliqué à la fois à la couverture et aux données vectorielles spatiales définies par des frontières.
5.5.2 Représentation utilisant des couvertures discrètes
Le résultat de la classification d’une zone est une légende qui peut être représentée comme un type de
couverture discrète, comme spécifié dans l’ISO 19123. La plage d’une couverture discrète est un ensemble de
valeurs d’attributs représenté comme un ensemble d’enregistrements ayant un schéma commun défini par le
système de classification. Le domaine consiste en une collection finie d’objets classifiés avec leurs positions
directes. L’objet géométrique et son enregistrement associé forment une paire de valeurs géométriques.
La classe CV_Coverage (voir l’ISO 19123) représente un type d’entité conforme au modèle général d’entité
issu de l’ISO 19109:2005, Figure 4. La classe CV_Coverage comporte trois attributs: domainExtent, rangeType
et commonPointRule. Dans des couvertures discrètes, il existe un rapport direct un à un où chaque CV_
GeometryValuePair est liée aux attributs correspondants. L’attribut domainExtent décrit l’étendue de la
couverture du domaine. Le type de données EX_Extent est défini dans l’ISO 19115. L’attribut rangeType
décrit la structure et la composition de l’enregistrement de données d’attribut. L’attribut rangeType utilise
le type de données RecordType, qui est défini dans l’ISO/TS 19103. Un RecordType est une métaclasse
qui décrit la structure d’un ensemble d’enregistrements. Un RecordType est composé d’une liste de paires
nom d’attribut/type de données. Une liste simple est la forme la plus commune de rangeType. L’attribut
commonPointRule identifie la procédure à utiliser pour l’évaluation de CV_Coverage à une position qui se
situe soit à la frontière entre des objets géométriques, soit dans les frontières de deux objets géométriques ou
plus se chevauchant, où les objets géométriques sont CV_DomainObjects ou CV_ValueObjects. Le type de
données CV_CommonPointRule est défini dans l’ISO 19123.
Une spécification du système de coordonnées de référence (CRS) qui référence les objets dans le domaine
est associée à une classe CV_Coverage. Le système de coordonnées de référence est défini dans l’ISO 19111.
Les classes CV_Domain et CV_AttributeValues sont également associées à la classe CV_Coverage. La classe
CV_Coverage relie l’ensemble des objets du domaine à l’ensemble d’enregistrements contenant les valeurs
d’attributs. Un CV_DomainObject peut être n’importe quel objet spatial, temporel ou spatio-temporel. Les
objets du domaine d’une CV_DiscreteSurfaceCoverage sont contraints d’être des instances de GM_Surface. Il
existe une instance de CV_AttributeValues (c’est-à-dire un enregistrement de données) pour chaque instance
de CV_DomainObject (c’est-à-dire chaque cellule de grille ou zone de polygone).
Voir Figure 4.
<>
GF_FeatureType
(à partir du modèle d’entité général)
<>
<>
CV_Coverage
(à partir de coverage core)
CoordinateReferenceSystem
+ domainExtent[1.*] : EX_Extent
0.*
+ rangeType : RecordType
+ commonPointRule : CV_CommonPointRule
+collection +collection
Plage
Domaine
1 +CRS 1.* +domainElement 0.* +rangeElement
<>
CV_DomainObject CV_AttributeValues
(à partir de coverage core) (à partir de coverage core)
SC_CRS
(à partir de référencement spatial par coordonnées)
+ values : Record
+spatialElement +temporalElement
SpatialComposition TemporalComposition
0.* 0.*
<>
TM_GeometricPrimitive
GM_Object (à partir des objets temporels)
(à partir de la racine géométrique)
Figure 4 — Éléments d’une couverture
CV_DiscreteCoverage est la sous-classe qui assigne le même enregistrement de valeurs d’attributs à
n’importe quelle position directe dans un seul CV_DomainObject dans son domaine. Chaque paire de valeurs
géométriques est composée d’un objet de domaine (c’est-à-dire une cellule de grille ou un polygone) et d’un
enregistrement des valeurs d’attributs. Voir Figure 5. Dans les couvertures discrètes, il existe un rapport direct
un à un où chaque CV_GeometryValuePair est composée d’un CV_DomainObject tel qu’un GM_Object et une
valeur CV_AttributeValues.
10 © ISO 2009 – Tous droits réservés
<>
CV_Coverage
(à partir de coverage core)
+ domainExtent[1.*] : EX_Extent
+ rangeType : RecordType
+ commonPointRule : CV_CommonPointRule
CV_DiscreteCoverage
(à partir des couvertures discrètes)
0.* +collection
CoverageFunction
1.* +element
CV_GeometryValuePair
(à partir de coverage core)
+ geometry : CV_DomainObject
+ value : Record
Figure 5 — Éléments d’une couverture discrète
Le type de couverture discrète est basé sur le type d’objet géométrique dans le domaine spatial. L’Article 6 de
l’ISO 19123:2005 identifie cinq types de couvertures discrètes ayant des géométries différentes:
— une couverture de points discrète, constituée d’un ensemble de points indépendants,
— une couverture de points de grille, constituée d’un ensemble de points de grille,
— une couverture de courbes discrète, constituée d’un ensemble de courbes,
— une couverture de surfaces discrète, constituée d’un ensemble de surfaces, en général des objets GM_
Surface (polygones), des rectangles isométriques (cellules de grille) ou des hexagones réguliers, mais
probablement des objets TIN ou des objets Polygone de Thiessen,
— une couverture de solides discrète, constituée d’un ensemble de volumes solides,
Un système de classification peut utiliser n’importe quel type d’objet géométrique dans son domaine spatial.
Il est possible d’établir des classificateurs pour identifier les valeurs de plage de n’importe quel type de
couverture discrète. Les couvertures discrètes qui fractionnent une zone sont traitées par une couverture de
surface discrète.
5.5.3 Couverture de surfaces discrète
Un système de classification qui utilise une couverture dont l’étendue fractionne une zone doit utiliser la
couverture de surfaces CV_Discrete conformément à l’ISO 19123:2005, 6.8. L’utilisation de GM_Surface
comme élément de géométrie est illustrée à la Figure 6.
CoverageFunction
CV_GeometryValuePair
CV_DiscreteCoverage (à partir de coverage core)
+collection +element
(à partir des couvertures discrètes)
0.* 1.*
+ geometry : CV_DomainObject
+ value : Record
CoverageFunction
CV_SurfaceValuePair
CV_DiscreteSurfaceCoverage
+collection +element
(à partir des couvertures discrètes)
(à partir des couvertures discrètes)
1.*
+ geometry : GM_Surface
Figure 6 — Couverture de surface discrète
Les surfaces constituant le domaine d’une couverture de surface discrète peuvent être mutuellement
exclusives et fractionner entièrement une zone ou peuvent être des sous-zones indépendantes. Le type de
géométrie de couverture utilisé dépend souvent de la méthode grâce à laquelle les données ont été recueillies.
Les phénomènes naturels tendent à avoir des frontières irrégulières et sont souvent représentés comme un
ensemble de polygones congruents. Certains phénomènes, types de données de capteurs ou résultats de
calculs tendent à être organisés en termes de structure de grille en mosaïque où le domaine est un ensemble
de rectangles congruents. Des sous-types de CV_DiscreteSurfaceCoverage pouvant être utilisés dans des
systèmes de classification sont illustrés à la Figure 7. Différentes structures de grilles en mosaïque peuvent
être utilisées. Une grille régulière avec des cellules de taille égale, ayant un ordre de séquence transversal
linéaire, est un type de grille quadrilatérale. Cependant, d’autres organisations de grilles peuvent être utilisées,
comme la grille «QuadTree» à taille de cellules variable ayant un ordre de séquence transversal de Morton.
CoverageFunction CV_SurfaceValuePair
CV_DiscreteSurfaceCoverage (à partir des couvertures discrètes)
1.*
(à partir des couvertures
+element
+collection + geometry : GM_Surface
discrètes)
+ value : Record
CL_ClassifiedSurface
+ tessellationGeometry [0.1] : CL_TessellationGeometry
+ sequenceRule : CV_SequenceType
<>
<> CV_SequenceType
(à partir d’une grille quadrilatérale)
CL_TessellationGeometry
+ linear
+ quadrilateral grid
+ boustrophedonic
+ hexagonal grid
+ CantorDiagonal
+ TIN
+ spiral
+ polygon network
+ Morton
+ Thiessen Polygon network
+ Hilbert
Figure 7 — Types de couvertures de surfaces discrètes
5.5.4 Attributs
Une paire de valeurs géométriques est composée d’un objet spatial, temporel ou spatio-temporel ainsi que
d’un enregistrement de va
...
Questions, Comments and Discussion
Ask us and Technical Secretary will try to provide an answer. You can facilitate discussion about the standard in here.
Loading comments...