ISO/DTS 25588

• Abstract:

  Specification of the set-up of a testing system and the test suite structure and test purposes, i.e. tests to be used to assess conformity to specification of the processes that implement an image-based tolling system.

• Status: Not Published
• ICS: 35.240.60 - IT applications in transport
• Technical Committee: ISO/TC 204 - Intelligent transport systems
• Drafting Committee: ISO/TC 204/WG 5 - Fee and toll collection
• Current Stage: 5020 - FDIS ballot initiated: 2 months. Proof sent to secretariat
• Start Date: 15-Jul-2026
• Completion Date: 15-Jul-2026

• Consolidates: FprCEN ISO/TS 25588 - Electronic fee collection - Image-based systems - Test suite structure and test purposes (ISO/DTS 25588:2026)
• Effective Date: 12-Feb-2026

Overview

ISO/DTS 25588:2026, developed by ISO/TC 204, specifies the requirements for the structure of test suites and the specification of test purposes to determine conformity in image-based electronic fee collection (EFC) systems. As image-based tolling systems become increasingly central to modern intelligent transport systems (ITS), effective and standardized testing is critical for ensuring accurate, reliable, and interoperable operation. This standard outlines laboratory and on-site test setups, interfaces, and events crucial for assessing core functions such as passage detection, vehicle identification, and vehicle classification.

Key Topics

• Test System Setup: ISO/DTS 25588 details both laboratory and operational (on-site) test configurations, including requirements for equipment such as cameras, licence plates, control workstations, and measurement devices.
• Test Suite Structure: The standard specifies a modular structure for test suites, including test scenarios that evaluate individual EFC components independently. This isolation ensures conformity to specification for each process before assessing overall system performance.
• Core Processes Tested:
  • Passage Detection: Validation of event generation when vehicles pass tolling points using technologies such as inductive loops, presence radars, and cameras.
  • Vehicle Identification: Evaluation of capabilities to correctly recognize licence plate numbers and related vehicle information, using optical character recognition (OCR) and automatic number plate recognition (ANPR).
  • Vehicle Classification: Assessment of classification algorithms that determine vehicle type, size, or axle count, often leveraging artificial intelligence (AI) or support vector machines (SVM).
• Data Format & Interfaces: Standardization of event generation, timestamping, interface requirements, and data formats ensures interoperability and facilitates result validation.

Applications

Practical Value for Stakeholders

Adoption of ISO/DTS 25588 supports a broad range of applications within the intelligent transportation ecosystem:

• Toll Service Providers: Ensure compliance with international specifications for image-based toll systems, improving reliability and harmonizing operations across regions.
• System Integrators: Utilize clearly defined test purposes and interfaces to select compatible components, facilitate integration, and expedite deployment.
• Regulators and Authorities: Rely on standardized test suites to verify system conformity, conduct audits, and maintain public trust in electronic tolling operations.
• Technology Developers: Reference test scenarios and data formats to design and validate new equipment (e.g., advanced ANPR cameras, AI classifiers).

Broader Use Cases

While the focus is on EFC/tolling, the framework can be adapted to related image-based intelligent transport system applications such as:

• Automated parking and access systems
• Vehicle compliance checks and roadside enforcement
• Traffic violation detection or monitoring

Related Standards

For comprehensive implementation and conformity in electronic fee collection and intelligent transport systems, the following ISO standards are relevant:

• ISO 17573-2: Electronic fee collection - System architecture for vehicle-related tolling - Vocabulary; referenced for terms and definitions.
• ISO/TR 25221: Classifies relevant characteristics of image-based systems and forms the basis for component-level testing.
• ISO/TS 37444 (future): Will define key performance indicators (KPIs) and numerical thresholds for EFC systems, integrating with the conformity assessment methods outlined here.

Conclusion

ISO/DTS 25588:2026 is essential for organizations seeking to validate compliance and performance in image-based EFC systems. By providing a standardized approach to test suite structure and test purposes, it promotes interoperability, transparency, and continual improvement across international electronic tolling operations and broader ITS applications. Implementers, regulators, and technology vendors all benefit from its clarity and actionable guidance.

FINAL DRAFT
Technical
Specification
ISO/TC 204
Electronic fee collection — Image-
Secretariat: ANSI
based systems —Test suite structure
Voting begins on:
and test purposes
2026-07-15
Perception de télépéage — Systèmes basés sur l'analyse d'images
Voting terminates on:
— Structure de la suite d’essais et objectifs des essais
2026-10-07
RECIPIENTS OF THIS DRAFT ARE INVITED TO SUBMIT,
WITH THEIR COMMENTS, NOTIFICATION OF ANY
RELEVANT PATENT RIGHTS OF WHICH THEY ARE AWARE
AND TO PROVIDE SUPPOR TING DOCUMENTATION.
IN ADDITION TO THEIR EVALUATION AS
BEING ACCEPTABLE FOR INDUSTRIAL, TECHNO­
ISO/CEN PARALLEL PROCESSING LOGICAL, COMMERCIAL AND USER PURPOSES, DRAFT
INTERNATIONAL STANDARDS MAY ON OCCASION HAVE
TO BE CONSIDERED IN THE LIGHT OF THEIR POTENTIAL
TO BECOME STAN DARDS TO WHICH REFERENCE MAY BE
MADE IN NATIONAL REGULATIONS.
Reference number
FINAL DRAFT
Technical
Specification
ISO/TC 204
Electronic fee collection — Image-
Secretariat: ANSI
based systems —Test suite structure
Voting begins on:
and test purposes
Perception de télépéage — Systèmes basés sur l'analyse d'images
Voting terminates on:
— Structure de la suite d’essais et objectifs des essais
RECIPIENTS OF THIS DRAFT ARE INVITED TO SUBMIT,
WITH THEIR COMMENTS, NOTIFICATION OF ANY
RELEVANT PATENT RIGHTS OF WHICH THEY ARE AWARE
AND TO PROVIDE SUPPOR TING DOCUMENTATION.
© ISO 2026
IN ADDITION TO THEIR EVALUATION AS
All rights reserved. Unless otherwise specified, or required in the context of its implementation, no part of this publication may
BEING ACCEPTABLE FOR INDUSTRIAL, TECHNO­
ISO/CEN PARALLEL PROCESSING
LOGICAL, COMMERCIAL AND USER PURPOSES, DRAFT
be reproduced or utilized otherwise in any form or by any means, electronic or mechanical, including photocopying, or posting on
INTERNATIONAL STANDARDS MAY ON OCCASION HAVE
the internet or an intranet, without prior written permission. Permission can be requested from either ISO at the address below
TO BE CONSIDERED IN THE LIGHT OF THEIR POTENTIAL
or ISO’s member body in the country of the requester.
TO BECOME STAN DARDS TO WHICH REFERENCE MAY BE
MADE IN NATIONAL REGULATIONS.
ISO copyright office
CP 401 • Ch. de Blandonnet 8
CH-1214 Vernier, Geneva
Phone: [phone]
Email: [e-mail]
Website: www.iso.org
Published in Switzerland Reference number
ii
Contents Page
Foreword .v
Introduction .vi
1 Scope . 1
2 Normative references . 1
3 Terms and definitions . 1
4 Abbreviated terms and symbols . 2
5 Processes and variables under test . 3
6 Test setup . 3
6.1 General setup requirements .3
6.2 Laboratory test setup .4
6.3 On site test setup .5
6.4 Passage detection specific setup .5
6.5 Vehicle identification specific setup .5
6.6 Classification specific setup .6
7 Test suite structure and interfaces . 6
7.1 General .6
7.2 Passage detection interfaces .7
7.2.1 General .7
7.2.2 Purpose .7
7.2.3 Event generation .7
7.2.4 Interface requirements . . .7
7.3 Vehicle identification interfaces . .8
7.4 Vehicle classification interfaces .8
7.4.1 General .8
7.4.2 Event generation .8
7.4.3 Interface requirements . . .8
8 Test purposes . 9
8.1 General .9
8.2 Naming .9
8.3 Format .9
8.4 Laboratory tests .10
8.4.1 General .10
8.4.2 Test setup and geometries .10
8.4.3 Static identification tests . 12
8.4.4 Dynamic identification tests . . 13
8.5 On site passage detection . 13
8.5.1 General . 13
8.5.2 Test setup . 13
8.5.3 Detection of true positives .14
8.5.4 Detection: false positives . 15
8.5.5 Detection: false negatives . 15
8.6 On site Vehicle identification . 15
8.6.1 Test setup . 15
8.6.2 Identification by licence plate reading: true positives . 15
8.6.3 Identification: False positives with licence plate reading .16
8.6.4 Identification:True negatives .17
8.7 Classification: vehicle classification .18
8.7.1 Test setup .18
8.7.2 Classification: True positives .19
8.7.3 Classification: True negatives .19
8.7.4 Classification: False negatives . 20

iii
Annex A (normative) Protocol Implementation eXtra Information for Test (PIXIT .21
Annex B (normative) Event data specification .24
Annex C (informative) Examples of real system implementations .25
Annex D (informative) Licence plate design and manufacturing considerations for image-based
vehicle identification .31
Bibliography .33

iv
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.
The procedures used to develop this document and those intended for its further maintenance are described
in the ISO/IEC Directives, Part 1. In particular, the different approval criteria needed for the different types
of ISO document should be noted. This document was drafted in accordance with the editorial rules of the
ISO/IEC Directives, Part 2 (see www.iso.org/directives).
ISO draws attention to the possibility that the implementation of this document may involve the use of (a)
patent(s). ISO takes no position concerning the evidence, validity or applicability of any claimed patent
rights in respect thereof. As of the date of publication of this document, ISO had not received notice of (a)
patent(s) which may be required to implement this document. However, implementers are cautioned that
this may not represent the latest information, which may be obtained from the patent database available at
www.iso.org/patents. ISO shall not be held responsible for identifying any or all such patent rights.
Any trade name used in this document is information given for the convenience of users and does not
constitute an endorsement.
For an explanation of the voluntary nature of standards, the meaning of ISO specific terms and expressions
related to conformity assessment, as well as information about ISO's adherence to the World Trade
Organization (WTO) principles in the Technical Barriers to Trade (TBT), see www.iso.org/iso/foreword.html.
This document was prepared by Technical Committee ISO/TC 204, Intelligent transport systems, in
collaboration with the European Committee for Standardization (CEN) Technical Committee CEN/TC 278,
Intelligent Transport Systems, in accordance with the Agreement on technical cooperation between ISO and
CEN (Vienna Agreement).
Any feedback or questions on this document should be directed to the user’s national standards body. A
complete listing of these bodies can be found at www.iso.org/members.html.

v
Introduction
[1]
ISO/TR 25221 identifies relevant characteristics of an image-based system and classifies these
characteristics in a number of processes that can be combined in various ways to perform services such
as electronic fee collection (EFC). The overall system performance or conformity to specifications depends
on the combination of the different processes (the architecture of the system) and the related business
processes, so it would be rather pointless to specify tests to measure it. However, test procedures can be
[1]
specified to determine conformity to specifications for the isolated processes that ISO/TR 25221 has
identified, as long as they can be accessed separately. This document specifies a component test suite for
image-based EFC systems.
Although this document is principally oriented at EFC systems, the test purposes herein specified are
considered general enough to be used to evaluate other image-based systems, such as:
— Parking
— Free-flow entry exit
— Gated parking
— Vehicle-based car park compliance checks and collection of evidence of non-compliance
— Vehicle-based roadside inspection or enforcement
— Traffic violation detection
— Roadside
— Vehicle-based
Key performance indicators (KPI) or numerical thresholds for parameters of EFC systems, including image-
[2]
based ones, are intended to be defined in a future edition of ISO/TS 37444 .

vi
FINAL DRAFT Technical Specification ISO/DTS 25588:2026(en)
Electronic fee collection — Image-based systems —Test suite
structure and test purposes
1 Scope
This document specifies the set-up of a testing system and the test suite structure and test purposes, i.e. tests
to assess conformity to specification for implemented processes of image-based electronic fee collection
(EFC) systems.
The test purposes specified in this document are solely for evaluating the behaviour of isolated processes in
an image-based EFC system.
The focus of the tests is related to the components and interfaces in the roadside system required for
fulfilling the needs of an image-based EFC system. Generic and overall tests related to the reliability and
qualitative capabilities of the complete charging point are outside of the scope of this document.
This document contains four annexes:
— Annex A, normative, that specifies the additional needed information that an implementation provides
to the tester to run tests;
— Annex B, normative, that specifies the format of the data to be produced for each test run;
— Annex C, informative, that collects a number of test purposes used to measure characteristics in
implemented systems;
— Annex D, informative, that collects licence plate design and manufacturing considerations.
2 Normative references
The following documents are referred to in the text in such a way that some or all of their content constitutes
requirements 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 17573-2, Electronic fee collection — System architecture for vehicle related tolling — Part 2: Vocabulary
3 Terms and definitions
For the purposes of this document, the terms and definitions given in ISO 17573-2 and the following apply.
ISO and IEC maintain terminology databases for use in standardization at the following addresses:
— IEC Electropedia: available at http:// www .electropedia .org/
— ISO Online browsing platform: available at http:// www .iso .org/ obp
3.1
performance test
testing that simulates the expected workload on an application to assess factors such as transaction speed,
user behaviour, and system stability under normal or peak conditions

3.2
identification
correct recognition of all characters contained in a licence plate and of the elements that identify the country
of registration of the vehicle
3.3
image-based EFC system identifier
device in the image-based EFC system that is in charge of a vehicle's licence plate identification (3.2)
3.4
reference space
volume where the image-based electronic fee collection (EFC) system is declared to be operating properly
3.5
actual distance
distance between the image-based EFC system identifier (3.3) and the vehicle's licence plate within which the
image-based EFC system is declared to be operating properly
4 Abbreviated terms and symbols
For the purposes of this document, the abbreviated terms in Table 1 apply.
Table 1 — Abbreviations
AD actual distance
AI artificial intelligence
ANPR automatic number plate recognition
EFC electronic fee collection
IBES image-based electronic fee collection system
LPN licence plate number
OCR optical character recognition
PD passage detection
PIXIT Protocol Implementation eXtra Information for Test
RS reference space
RSE roadside equipment
RSW reference space width
SUT system under test
SVM support vector machine
TCLS toll charger local system
VI vehicle identification
VC vehicle classification
For the purposes of this document, the symbols in Table 2 apply.
Table 2 — Symbols
Symbol Meaning
DET detection of true positive vehicles
CDI classification rate when classification is independent of detection and identification
CIC classification rate when classification happens after identification
CLI classification rate dependent on licence plate reading
CNR classification true negative rate
DFP detection of false positives rate

TTabablele 2 2 ((ccoonnttiinnueuedd))
Symbol Meaning
ICI identification rate when classification precedes identification
IDI identification rate when detection precedes identification
LSI laboratory static identification
5 Processes and variables under test
[3]
Table 3, derived from Table 5 in ISO/TR 25221:2025 , shows the variables and processes that may be
subject to conformance test both in laboratory and in the field. The latter tests are at the frontier between
conformance and performance tests.
Table 3 — Process and variables subject to tests
Laboratory test Test in field
Process Variable
Testability Conditions Testability Conditions
Passage Detection rate No Yes Additional trusted
detection detection system
available
Detection of false No Yes Additional trusted
positives rate detection system
available
Detection of false No Yes Additional trusted
negatives rate detection system
available
Vehicle Identification rate Yes Optical charac- Yes OCR separately
identification when detection ter recognition testable
precedes identifica- (OCR) separately
tion testable
Identification rate Yes OCR separately Yes OCR separately
when classification testable testable
precedes identifica-
tion
Classification Classification rate Yes Classification Yes Additional trusted
when independent system separate- detection system
of detection and ly testable available
identification
Classification rate Yes Classification Yes Additional trusted
when classification system separate- identification sys-
happens after iden- ly testable tem available
tification
6 Test setup
6.1 General setup requirements
Subclauses 6.1 to 6.6 describe the general setup for laboratory and on site tests. Further details regarding
any specific tests are specified, if necessary, in the relative paragraphs.
Some assumptions are taken about licence plates, which shall be noted and reported when performing tests.
These include, but are not limited to:
— Modern cameras use both natural light and IR images in their detection process.
— All plates are retro-reflective. This is important to locate the plate on a vehicle. The retro-reflectivity
provides a high contrast rectangle to find.

— Plate colour, fonts, images, labels are regulated to ensure the plate and its characters are clearly
identifiable.
— Plates are fitted to vehicles in a predictable manner in a predictable area.
— Frames and screws have a negative impact on ANPR, specifically obscuring the retro-reflectivity of the
plate.
— The measurement does not deal with plate-tampering to confuse detection, though the test as written
can be used for that purpose as well.
— Poor plate material quality and tooling have an impact on ANPR performance as plates age. The main
issues are:
— Delamination.
— Character print decal impacts sharpness of character edges.
The system under test (SUT) shall expose a fully operational replica of the roadside equipment (RSE) and
any centralized correlation component used in operation. When correlation among multiple RSE sources is
centralized, the same correlation logic shall be part of the SUT during tests to surface possible end-to-end
failure modes (e.g. race conditions, late event merging).
Specific requirements that the SUT shall provide for testing purposes are detailed in Annex A.
6.2 Laboratory test setup
Laboratory tests aim at verifying:
— the ability of the system to correctly acquire and recognize license plates under controlled lighting,
geometry, motion and occlusion conditions;
— the robustness of the ANPR process under variations of perspective, illumination, plate position and
simultaneous presence of multiple plates within the reference space (RS).
The tests shall be conducted in a controlled indoor environment, providing:
— controlled ambient light levels (0 lux to 10,000 lux depending on the test);
— free space to position plates and cameras at required geometries;
— mechanical equipment enabling precise placement and rotation of plates.
Laboratory instrumentation includes:
— A set of licence plates manufactured with materials compliant with current regulatory prescriptions (as
per the applicable highway code). Plates may be of any of the following types:
— rear plates for motor vehicles;
— front plates for motor vehicles;
— rear plates for motorcycles;
— plates for mopeds.
— An RS with the following characteristics:
— Adjustable incandescent (e.g. halogen) lamp with illuminance levels as required by the tests and
horizontal incidence angle of 13° ± 1° relative to plate axis.

— Mechanical plate positioning system enabling rotation and placement of plates at multiple angles
and positions within the RS.
— For some specific requirements, a motion simulation system simulating relative motion between camera
and plate, equipped with optoelectronic trigger sensor to generate the capture event.
— A speed measurement device (e.g. radar or laser).
— A control workstation.
— Measuring equipment: surveyor’s measuring tape and steel tape measure.
— Test license plates (assorted series).
6.3 On site test setup
The SUT shall provide a fully functional replica of the operational system in an area capable of accommodating
transits by different vehicle classes.
Depending on the test purpose of the specific test, if correlation between outputs from multiple RSE units
is centralized, the SUT shall replicate the centralized system to demonstrate both the provenance of each
result and any potential weaknesses in the correlation process.
Unless stated otherwise, the on-site test fleet shall include at least: 1 motorcycle, 1 car, 1 light truck, 1 heavy
or articulated truck (if allowed by the site), optionally 1 bicycle and 1 pedestrian for false-positive checks.
As far as environmental conditions are concerned, daylight sessions are mandatory; dusk or night IR
sessions are recommended. Disturbances such as headlamps, reflections and wet surface should be included
when feasible.
For distances among vehicles and vehicles sequences, use separated flows (≈5 m gaps) and tailgating
scenarios (≤1 m, safety-permitting) at controlled speeds (e.g. 10–30 km/h on test track; higher speeds if
declared by the manufacturer and allowed by the site).
The geometry of the testing site shall reflect a realistic operational deployment (gantry, portal, side-
mounted, tripod, mobile, on-board) and be declared in the Protocol Implementation eXtra Information for
Test (PIXIT, see Annex A).
NOTE Realistic operational deployment" can be representative of an installation instance, or the most challenging
scenario of all the installation instance of a system.
Where the SUT relies on multiple sensors for detection and classification, each source shall be logged
separately, and the correlation algorithm shall be briefly described to the tester.
6.4 Passage detection specific setup
Based on the setup specified in 6.3, if the vehicle detection process is based on an RSE output, and the output
consists of more than a simple ON/OFF signal , the SUT shall provide access to an interface (hardware or
software) to assess the occurred detection (see Annex A).
Detection events shall be produced in a standardized format and time-stamped to detect generation delays
(see Annex B).
The aforementioned interface shall produce a single identifiable output for each passage.
6.5 Vehicle identification specific setup
Based on the setup specified in 6.3, if the vehicle identification process is based on a RSE output, the SUT
shall provide access to an hardware or software interface to visualize the identification of the vehicle.
If identification relies on RSE output, expose an interface to visualize identification results (e.g. licence
plate number (LPN) string, nationality, region, confidence score). When the SUT can identify other physical

objects, provide a reference table describing all identifiable objects. One identifiable output per passage is
required.
Prior to testing, it shall be verified that the plate is fully contained in the declared area across a calibration
grid and typical angles; simulate low-speed motion to confirm absence of excessive motion blur.
6.6 Classification specific setup
Based on the setup specified in paragraph 6.3, if the vehicle classification process is based on an RSE output,
the SUT shall provide access to an hardware or software interface to visualize the occurred classification of
the vehicle.
The aforementioned interface shall produce a single identifiable output for each passage.
7 Test suite structure and interfaces
7.1 General
Figure 1 shows the physical and functional high-level system architecture for an image-based EFC system.
Focus is on the interfaces that are relevant for the scope of this document.
Figure 1 — High-level system architecture for an image-based EFC system
The image-based EFC system (IBES) shall provide a reference of the time zone when in operation and it shall
provide a synchronization reference status available during the whole test process.

7.2 Passage detection interfaces
7.2.1 General
The passage detection interfaces consist of the three interfaces between the following physical objects:
— Inductive loop and Toll Charger Local System (TCLS). The inductive loop sends a message to the TCLS
whenever the inductive loop detects a vehicle. Based on the configuration of the loop, the loop may also
transmit data about the size of the vehicle.
— Vehicle presence radar and the TCLS. The radar sends a message to the TCLS whenever an object
with a size indicating that the object detected is a vehicle. The radar may also send data describing the
physical parameters of the physical object detected, see 7.4.
— Vehicle presence camera and the TCLS. The camera sends a message to the TCLS whenever a camera
has detected an object in the camera view that indicates that a vehicle is present in the toll lane.
7.2.2 Purpose
To define the interface specifications for detecting the passage of vehicles through tolling points, ensuring
accurate and reliable toll collection operations.
The SUT separation and distinction capabilities shall be assessed to avoid any possible fraudulent use of the
TCLS.
7.2.3 Event generation
Upon vehicle detection, the SUT shall generate an event containing:
— a timestamp,
— an optional lane identifier,
— the detection status (e.g success, failure),
— the utilized detection method.
If the detection process is more complex than a single ON/OFF signal, and a sequence of detection events is
identifiable, the timestamp generated shall indicate both the start and the stop event times.
The specification of the event data is specified in Annex B.
7.2.4 Interface requirements
7.2.4.1 General
The SUT may utilize sensors (e.g. loop detectors, infrared sensors, LIDAR, high-speed cameras) to detect
vehicles moving with different speed and across various lanes.
The speed ranges used in tests shall be compliant to those declared on the road signage, or those defined by
the national law.
7.2.4.2 Data format and transmission
Detection events shall be transmitted in a format known and understandable by the tester to the central
processing unit in real-time. The type and the characteristics of that format would be SUT-specific and are
not specified by this document.
When running tests, the event generation, among other data, shall be identified by a timestamp (specified in
Annex B), to detect possible delay between the time period of the detection process and the event generation.

7.2.4.3 Error handling and redundancy
Missed detections shall be logged and flagged for review.
7.3 Vehicle identification interfaces
In image-based EFC systems, a vehicle identification interface is identified between the ANPR camera and
the TCLS. ANPR cameras may be used to capture both front and rear licence plates of the vehicle. Cameras
may also capture pictures of the complete vehicle (front and rear), to support verification of the vehicle
identification by additional detected information, e.g. the make and the colour of the vehicle, and to compare
them to the officially registered information that is associated to the licence plate.
7.4 Vehicle classification interfaces
7.4.1 General
Vehicle classification interfaces are identified between the following physical objects:
— Vehicle classification camera and TCLS. The camera sends pictures of the vehicle to the TCLS and
the TCLS uses the pictures possibly supported by artificial intelligence (AI) to classify the vehicle (e.g.
determining its size and number of axles). This enables the toll charger (TC) to verify that the vehicle
detected data correspond to those officially registered.
— Vehicle classification radar or lidar and TCLS. The radar or lidar sends vehicle detected data to the
TCLS and the TCLS uses that data (possibly supported by AI) to classify the vehicle (e.g. determining
its size and number of axles). This enables the TC to verify that the vehicle detected data correspond to
those officially registered.
7.4.2 Event generation
Upon vehicle detection, the SUT shall generate an event containing:
— a timestamp,
— a lane identifier,
— the detection status (e.g. success, failure),
— the utilized detection method.
If specific characteristics are used for the classification method (e.g. number of axles) the information shall
be provided in the generated event.
If a national classification scheme exists, and it is different from the one used by the TCLS, both classifications
shall be provided in the generated event, each of them associated with an accuracy evaluation and the
method utilized (e.g. the national scheme can be derived by an algorithm applied to the TCLS classification).
7.4.3 Interface requirements
7.4.3.1 Classification mechanism
The classification mechanism shall be described by the SUT.
In case of an AI, support vector machine (SVM) or neural network detection process, the basic logic behind
the classification process shall also be declared and described.
NOTE If the classification process is based on national characteristics (e.g. specific markings fixed to the vehicle,
sequence of letters in the licence plate), the TCLS is generally not able to operate on foreign vehicles.

7.4.3.2 Data format and transmission
Classification events shall be transmitted in a specified format to the central processing unit in real-time, as
a single event or as a part of the picture metadata.
The test generated event shall be identified by a timestamp, to detect possible delays between the time
period of the classification process and the event generation. The format of the test generated events is
specified in Annex B.
7.4.3.3 Error handling and redundancy
Missed classification shall be logged and flagged for review.
In case of a classification score, an acceptability threshold shall be declared by the manufacturer, and all the
transits below the acceptability threshold shall be flagged for review.
8 Test purposes
8.1 General
Test purposes are specified in this document for the three capabilities of an image-based EFC system to
detect the passage of vehicles, to identify vehicles, and to classify vehicles.
These three capabilities are supposedly be tested independently, i.e. irrespective of the image-based EFC
system architecture.
The abbreviations in Table 1 will be used for naming the test purposes, see 8.2.
8.2 Naming
Test purposes are named in the following as XXYYYZZZ, where:
— [XX] indicates the process name, as per Table 1;
— [YYY] indicates the variable name, as per Table 2;
— [ZZZ] indicates the sequence number of the test purpose.
8.3 Format
Test purposes are specified in a table format, as shown by the proforma in Table 4, where the words in
boldface are entry names.
Table 4 — Test purpose proforma
Test Purpose name Test purpose description
Initial conditions Initial conditions
Repetitions Number of repetitions of the test purpose
Step sequence Step description
Sequence number Description of the step
Expected results Results expected. They may be expressed in terms of, e.g., number or percentage of true
positive results.
8.4 Laboratory tests
8.4.1 General
The following tests are intended to be performed in a closed laboratory with the characteristics and setups
as specified in 6.2. Laboratory tests are only aimed at identifying the vehicle by correctly reading its licence
plate [see Identification (of a licence plate)].
All laboratory tests shall be conducted with:
— camera aligned frontally to the licence plate (a limited lateral offset is allowed to cater for a lane width);
— installation height = manufacturer declared height or max laboratory height;
— licence plate distance configured to maintain the declared actual distance (AD).
8.4.2 Test setup and geometries
The following geometric parameters shall be declared by the manufacturer and used for testing (see
Figure 2):
— H: maximum height of the camera,
— AD: actual distance between camera and licence plate,
— LO: lateral offset between the camera and the licence plate,
— GD: ground distance between the licence plate and the plane perpendicular to the licence plate where the
camera is positioned.
Figure 2 — Geometric parameters for laboratory tests
The same geometrical terms apply for the evaluation of:
— RSW (Reference Space Width at 0 lux),
— Identification depth, the maximum GD for which valid identification is verified,
— Limit positions used in static recognition tests. These are the points, in the coordinates (x,y,z) as shown
in Figure 2, where the plate is positioned:
— The centre lane position at ground level (GD,0,0)
— The centre lane position at half of the maximum height of the camera (GD, 0, H/2)
— The centre lane position at maximum height of the camera (GD, 0, H)
— The right displacement from the centre lane at ground level (GD, LO, 0)
— The right displacement from the centre lane at half of the maximum height of the camera (GD, LO,
H/2)
— The right displacement from the centre lane at maximum height of the camera (GD, LO, H)
— The left displacement from the centre lane at ground level (GD, -LO, 0)
— The left displacement from the centre lane at half of the maximum height of the camera (GD, -LO,
H/2)
— The left displacement from the centre lane at maximum height of the camera (GD, LO, -H)

8.4.3 Static identification tests
Static identification tests, specified in Table 5, Table 6 and Table 7, are aimed at measuring the ability of
identifying static licence plates in different light conditions.
Table 5 — TP VILSI001: Identification of static licence plate in the laboratory in 0 lux illumination
conditions
VILSI001 Identification of vehicles in 0 lux light conditions
Initial conditions Position each licence plate at the nine limit points within the RS.
Consider three azimuth displacements for the licence plate: 0°, +αmax, -αmax.
Set illuminance to 0 lux.
Repetitions Repeat each step for each limit point and each azimuth angle for 100 times, for a total
of 2700 identifications
Step sequence Step description
1 Position the licence plate at one limit point at one azimuth displacement
2 Identify the licence plate and save the result
Expected results For each position and each azimuth displacement, the percentage of valid identifica-
tions shall be better or equal to the expected value.
Table 6 — TP VILSI002: Identification of static licence plate in the laboratory in 10000 lux illumination
conditions
VILSI002 Identification of vehicles in 10000 lux light conditions
Initial conditions Position each licence plate at the nine limit points within the RS.
Consider three azimuth displacements for the licence plate: 0°, +αmax, -αmax.
Set illuminance to 10000 lux.
Repetitions Repeat each step for each limit point and each azimuth angle for 100 times, for a total
of 2700 identifications
Step sequence Step description
1 Position the licence plate at one limit point at one azimuth displacement .
2 Identify the licence plate and save the result
Expected results For each position and each azimuth displacement, the percentage of valid identifica-
tions shall be better or equal to the expected value.
Table 7 — TP VILSI003: Identification of static licence plate in the laboratory in the presence of
shadows
VILSI003 Identification of vehicles in the presence of shadows
Initial conditions Position each licence plate at the nine limit points within the RS.
Consider three azimuth displacements for the licence plate: 0°, +αmax, -αmax.
Set two difference illuminances for two sectors of the licence plate, one to 10000 lux
and another one to 100 lux, in such a way to divide the licence plate into two different-
ly lighten areas. Prepare two possibilities: one where the licence plate is differently
shadowed vertically and one where the licence plate is differently shadowed horizon-
tally.
Repetitions Repeat each step for each limit point, each azimuth angle and each for every mode of
shadowing for 100 times, for a total of 5400 identifications
Step sequence Step description
1 Position the licence plate at one limit point at one azimuth displacement with one mode
of shadowing.
2 Identify the licence
...

ISO/TC 204
ISO/CD TS 25588(en)
Secretariat: ANSI
Date: 2026-06-30
Electronic fee collection — Image-based systems —Test suite
structure and test purposes
Perception de télépéage — Systèmes basés sur l'analyse d'images — Structure de la suite d’essais et objectifs
des essais
TTThhhiiisss d d drrraftaftaft i i isss s s suuubbbmmmiiittttttededed t t tooo  a pa pa parararallel vallel vallel vooottte e e iiinnn I I ISSSOOO,,, C C CEEENNN.

All rights reserved. Unless otherwise specified, or required in the context of its implementation, no part of this publication
may be reproduced or utilized otherwise in any form or by any means, electronic or mechanical, including photocopying,
or posting on the internet or an intranet, without prior written permission. Permission can be requested from either ISO
at the address below or ISO’s member body in the country of the requester.
ISO copyright office
CP 401 • Ch. de Blandonnet 8
CH-1214 Vernier, Geneva
Phone: + 41 22 749 01 11
E-mail: [e-mail]
Website: www.iso.org
Published in Switzerland
ii
Contents
Foreword . iv
Introduction . v
1 Scope . 1
2 Normative references . 1
3 Terms and definitions . 1
4 Abbreviated terms and symbols . 2
5 Processes and variables under test . 3
6 Test setup. 4
6.1 General setup requirements . 4
6.2 Laboratory test setup . 4
6.3 On site test setup . 5
6.4 Passage detection specific setup . 6
6.5 Vehicle identification specific setup . 6
6.6 Classification specific setup . 6
7 Test suite structure and interfaces . 6
7.1 General. 6
7.2 Passage detection interfaces . 8
7.3 Vehicle identification interfaces . 9
7.4 Vehicle classification interfaces . 10
8 Test purposes . 11
8.1 General. 11
8.2 Naming . 11
8.3 Format . 11
8.4 Laboratory tests . 11
8.5 On site passage detection . 17
8.6 On site Vehicle identification . 19
8.7 Classification: vehicle classification . 22
Annex A (normative) Protocol Implementation eXtra Information for Test (PIXIT . 25
Annex B (normative) Event data specification . 28
Annex C (informative) Examples of real system implementations . 29
Annex D (informative) Licence plate design and manufacturing considerations for image-based
vehicle identification . 35
Bibliography . 37

iii
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.
The procedures used to develop this document and those intended for its further maintenance are described
in the ISO/IEC Directives, Part 1. In particular, the different approval criteria needed for the different types of
ISO document should be noted. This document was drafted in accordance with the editorial rules of the
ISO/IEC Directives, Part 2 (see www.iso.org/directives).
ISO draws attention to the possibility that the implementation of this document may involve the use of (a)
patent(s). ISO takes no position concerning the evidence, validity or applicability of any claimed patent rights
in respect thereof. As of the date of publication of this document, ISO [had/had not] received notice of (a)
patent(s) which may be required to implement this document. However, implementers are cautioned that this
may not represent the latest information, which may be obtained from the patent database available at
www.iso.org/patents. ISO shall not be held responsible for identifying any or all such patent rights.
Any trade name used in this document is information given for the convenience of users and does not
constitute an endorsement.
For an explanation of the voluntary nature of standards, the meaning of ISO specific terms and expressions
related to conformity assessment, as well as information about ISO's adherence to the World Trade
Organization (WTO) principles in the Technical Barriers to Trade (TBT), see www.iso.org/iso/foreword.html.
This document was prepared by Technical Committee ISO/204, Intelligent transport systems.TC 204,
Intelligent transport systems, in collaboration with the European Committee for Standardization (CEN)
Technical Committee CEN/TC 278, Intelligent Transport Systems, in accordance with the Agreement on
technical cooperation between ISO and CEN (Vienna Agreement).
Any feedback or questions on this document should be directed to the user’s national standards body. A
complete listing of these bodies can be found at www.iso.org/members.html.
iv
Introduction
ISO/TR 25221 identifies relevant characteristics of an image-based system and classifies these characteristics
in a number of processes that can be combined in various ways to perform services such as electronic fee
collection (EFC). The overall system performance or conformity to specifications depends on the combination
of the different processes (the architecture of the system) and the related business processes, so it would be
rather pointless to specify tests to measure it. However, test procedures can be specified to determine
conformity to specifications for the isolated processes that ISO/TR 25221 has identified, as long as they can
be accessed separately. This document specifies a component test suite for image-based EFC systems.
Although this document is principally oriented at EFC systems, the test purposes herein specified are
considered general enough to be used to evaluate other image-based systems, such as:
— Parking
— Free-flow entry exit
— Gated parking
— Vehicle-based car park compliance checks and collection of evidence of non-compliance
— Vehicle-based roadside inspection or enforcement
— Traffic violation detection
— Roadside
— Vehicle-based
Key performance indicators (KPI) or numerical thresholds for parameters of EFC systems, including image-
based ones, are intended to be defined in a future edition of ISO/TS 37444.
Field Code Changed
v
Electronic fee collection — Image-based systems —Test suite
structure and test purposes
1 Scope
This document specifies the set-up of a testing system and the test suite structure and test purposes, i.e. tests
to assess conformity to specification for implemented processes of image-based electronic fee collection (EFC)
systems.
The test purposes specified in this document are solely for evaluating the behaviour of isolated processes in
an image-based EFC system.
The focus of the tests is related to the components and interfaces in the roadside system required for fulfilling
the needs of an image-based EFC system. Generic and overall tests related to the reliability and qualitative
capabilities of the complete charging point are outside of the scope of this document.
This document contains four annexes:
— Annex A, normative, that specifies the additional needed information that an implementation provides to
the tester to run tests;
— Annex B, normative, that specifies the format of the data to be produced for each test run;
— Annex C, informative, that collects a number of test purposes used to measure characteristics in
implemented systems;
— Annex D, informative, that collects licence plate design and manufacturing considerations.
2 Normative references
The following documents are referred to in the text in such a way that some or all of their content constitutes
requirements 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 17573-2, Electronic fee collection — System architecture for vehicle related tolling — Part 2: Vocabulary
3 Terms and definitions
For the purposes of this document, the terms and definitions given in ISO 17573-2 and the following apply.
ISO and IEC maintain terminologicalterminology databases for use in standardization at the following
addresses:
— IEC Electropedia: available at http://www.electropedia.org/
— ISO Online browsing platform: available at http://www.iso.org/obp
In addition to those definitions, the following definitions apply specifically to this document.
3.1
performance test
testing that simulates the expected workload on an application to assess factors such as transaction speed,
user behaviour, and system stability under normal or peak conditions
3.2
identification (of a licence plate)
correct recognition of all characters contained in a licence plate and of the elements that identify the country
of registration of the vehicle
3.3
image-based EFC system identifier
device in the image-based EFC system that is in charge of a vehicle's licence plate identification identification
(of a licence plate) (3.2)
3.4
reference space
volume where the image-based electronic fee collection (EFC) system is declared to be operating properly
3.5
actual distance
distance between the image-based EFC system identifier (3.3) and the vehicle's licence plate within which the
image-based EFC system is declared to be operating properly
4 Abbreviated terms and symbols
For the purposes of this document, the abbreviated terms in Table 1 apply.
Table 1 — Abbreviations
AD actual distance
AI artificial intelligence
ANPR automatic number plate recognition
EFC electronic fee collection
IBES image-based electronic fee collection system
LPN licence plate number
OCR optical character recognition
PD passage detection
PIXIT Protocol Implementation eXtra Information for Test
RS reference space
RSE roadside equipment
RSW reference space width
SUT system under test
SVM support vector machine
TCLS toll charger local system
VI vehicle identification
VC vehicle classification
For the purposes of this document, the abbreviated termssymbols in Table 2 apply.
Table 2 — Symbols
Symbol Meaning
DET detection of true positive vehicles
CDI classification rate when classification is independent of detection and identification
CIC classification rate when classification happens after identification
CLI classification rate dependent on licence plate reading
CNR classification true negative rate
DFP detection of false positives rate
ICI identification rate when classification precedes identification
IDI identification rate when detection precedes identification
LSI laboratory static identification
5 Processes and variables under test
Table 3, derived from Table 5 in ISO/TR 25221:2025, shows the variables and processes that may be subject
to conformance test both in laboratory and in the field. The latter tests are at the frontier between
conformance and performance tests.
Table 3 — Process and variables subject to tests
Laboratory test Test in field
Process Variable
Testability Conditions Testability Conditions
Passage Detection rate No Yes Additional
detection trusted detection
system available
Detection of false No Yes Additional
positives rate trusted detection
system available
Detection of false No Yes Additional
negatives rate trusted detection
system available
Vehicle Identification rate Yes Optical Yes OCR separately
identification when detection character testable
precedes recognition
identification (OCR)
separately
testable
Identification rate Yes OCR separately Yes OCR separately
when classification testable testable
precedes
identification
Classification Classification rate Yes Classification Yes Additional
when independent system trusted detection
of detection and separately system available
identification testable
Classification rate Yes Classification Yes Additional
when classification system trusted
Laboratory test Test in field
Process Variable
Testability Conditions Testability Conditions
happens after separately identification
identification testable system available
6 Test setup
6.1 General setup requirements
Subclauses 6.1 to 6.6 describe the general setup for laboratory and on site tests. Further details regarding any
specific tests are specified, if necessary, in the relative paragraphs.
Some assumptions are taken about licence plates, which shall be noted and reported when performing tests.
These include, but are not limited to:
— Modern cameras use both natural light and IR images in their detection process.
— All plates are retro-reflective. This is important to locate the plate on a vehicle. The retro-reflectivity
provides a high contrast rectangle to find.
— Plate colour, fonts, images, labels are regulated to ensure the plate and its characters are clearly
identifiable.
— Plates are fitted to vehicles in a predictable manner in a predictable area.
— Frames and screws have a negative impact on ANPR, specifically obscuring the retro-reflectivity of the
plate.
— The measurement does not deal with plate-tampering to confuse detection, though the test as written can
be used for that purpose as well.
— Poor plate material quality and tooling have an impact on ANPR performance as plates age. The main
issues are:
— Delamination.
— Character print decal impacts sharpness of character edges.
The system under test (SUT) shall expose a fully operational replica of the roadside equipment (RSE) and any
centralized correlation component used in operation. When correlation among multiple RSE sources is
centralized, the same correlation logic shall be part of the SUT during tests to surface possible end-to-end
failure modes (e.g. race conditions, late event merging).
Specific requirements that the SUT shall provide for testing purposes are detailed in Annex A.
6.2 Laboratory test setup
Laboratory tests aim at verifying:
— the ability of the system to correctly acquire and recognize license plates under controlled lighting,
geometry, motion and occlusion conditions;
— the robustness of the ANPR process under variations of perspective, illumination, plate position and
simultaneous presence of multiple plates within the reference space (RS).
The tests shall be conducted in a controlled indoor environment, providing:
— controlled ambient light levels (0 lux to 10,000 lux depending on the test);
— free space to position plates and cameras at required geometries;
— mechanical equipment enabling precise placement and rotation of plates.
Laboratory instrumentation includes:
— A set of licence plates manufactured with materials compliant with current regulatory prescriptions (as
per the applicable highway code). Plates may be of any of the following types:
— rear plates for motor vehicles;
— front plates for motor vehicles;
— rear plates for motorcycles;
— plates for mopeds.
— An RS with the following characteristics:
— Adjustable incandescent (e.g. halogen) lamp with illuminance levels as required by the tests and
horizontal incidence angle of 13° ± 1° relative to plate axis.
— Mechanical plate positioning system enabling rotation and placement of plates at multiple angles and
positions within the RS.
— For some specific requirements, a motion simulation system simulating relative motion between camera
and plate, equipped with optoelectronic trigger sensor to generate the capture event.
— A speed measurement device (e.g. radar or laser).
— A control workstation.
— Measuring equipment: surveyor’s measuring tape and steel tape measure.
— Test license plates (assorted series).
6.3 On site test setup
The SUT shall provide a fully functional replica of the operational system in an area capable of accommodating
transits by different vehicle classes.
Depending on the test purpose of the specific test, if correlation between outputs from multiple RSE units is
centralized, the SUT shall replicate the centralized system to demonstrate both the provenance of each result
and any potential weaknesses in the correlation process.
Unless stated otherwise, the on-site test fleet shall include at least: 1 motorcycle, 1 car, 1 light truck, 1 heavy
or articulated truck (if allowed by the site), optionally 1 bicycle and 1 pedestrian for false-positive checks.
As far as environmental conditions are concerned, daylight sessions are mandatory; dusk or night IR sessions
are recommended. Disturbances such as headlamps, reflections and wet surface should be included when
feasible.
For distances among vehicles and vehicles sequences, use separated flows (≈5 m gaps) and tailgating scenarios
(≤1 m, safety-permitting) at controlled speeds (e.g. 10–30 km/h on test track; higher speeds if declared by the
manufacturer and allowed by the site).
The geometry of the testing site shall reflect a realistic operational deployment (gantry, portal, side-mounted,
tripod, mobile, on-board) and be declared in the Protocol Implementation eXtra Information for Test (PIXIT,
see Annex A ).).
NOTE realisticRealistic operational deployment" can be representative of an installation instance, or the most
challenging scenario of all the installation instance of a system.
Where the SUT relies on multiple sensors for detection and classification, each source shall be logged
separately, and the correlation algorithm shall be briefly described to the tester.
6.4 Passage detection specific setup
Based on the setup specified in 6.3, if the vehicle detection process is based on an RSE output, and the output
consists of more than a simple ON/OFF signal , the SUT shall provide access to an interface (hardware or
software) to assess the occurred detection (see Annex A).
Detection events shall be produced in a standardized format and time-stamped to detect generation delays
(see Annex B).
The aforementioned interface shall produce a single identifiable output for each passage.
6.5 Vehicle identification specific setup
Based on the setup specified in 6.3, if the vehicle identification process is based on a RSE output, the SUT shall
provide access to an hardware or software interface to visualize the identification of the vehicle.
If identification relies on RSE output, expose an interface to visualize identification results (e.g. licence plate
number (LPN) string, nationality, region, confidence score). When the SUT can identify other physical objects,
provide a reference table describing all identifiable objects. One identifiable output per passage is required.
Prior to testing, it shall be verified that the plate is fully contained in the declared area across a calibration grid
and typical angles; simulate low-speed motion to confirm absence of excessive motion blur.
6.6 Classification specific setup
Based on the setup specified in paragraph 6.3, if the vehicle classification process is based on an RSE output,
the SUT shall provide access to an hardware or software interface to visualize the occurred classification of
the vehicle.
The aforementioned interface shall produce a single identifiable output for each passage.
7 Test suite structure and interfaces
7.1 General
Figure 1 shows the physical and functional high-level system architecture for an image-based EFC system.
Focus is on the interfaces that are relevant for the scope of this document.
Figure 1 — High-level system architecture for an image-based EFC system
The image-based EFC system (IBES) shall provide a reference of the time zone when in operation and it shall
provide a synchronization reference status available during the whole test process.
7.2 Passage detection interfaces
7.2.1 General
The passage detection interfaces consist of the three interfaces between the following physical objects:
— Inductive loop and Toll Charger Local System (TCLS). The inductive loop sends a message to the TCLS
whenever the inductive loop detects a vehicle. Based on the configuration of the loop, the loop may also
transmit data about the size of the vehicle.
— Vehicle presence radar and the TCLS. The radar sends a message to the TCLS whenever an object with
a size indicating that the object detected is a vehicle. The radar may also send data describing the physical
parameters of the physical object detected, see 7.4.
— Vehicle presence camera and the TCLS. The camera sends a message to the TCLS whenever a camera
has detected an object in the camera view that indicates that a vehicle is present in the toll lane.
7.2.2 Purpose
To define the interface specifications for detecting the passage of vehicles through tolling points, ensuring
accurate and reliable toll collection operations.
The SUT separation and distinction capabilities shall be assessed to avoid any possible fraudulent use of the
TCLS.
7.2.3 Event generation
Upon vehicle detection, the SUT shall generate an event containing:
— a timestamp,
— an optional lane identifier,
— the detection status (e.g success, failure),
— the utilized detection method.
If the detection process is more complex than a single ON/OFF signal, and a sequence of detection events is
identifiable, the timestamp generated shall indicate both the start and the stop event times.
The specification of the event data is specified in Annex B.
7.2.4 Interface requirements
7.2.4.1 General
The SUT may utilize sensors (e.g. loop detectors, infrared sensors, LIDAR, high-speed cameras) to detect
vehicles moving with different speed and across various lanes.
The speed ranges used in tests shall be compliant to those declared on the road signage, or those defined by
the national law.
7.2.4.2 Data format and transmission
Detection events shall be transmitted in a format known and understandable by the tester to the central
processing unit in real-time. The type and the characteristics of that format would be SUT-specific and are not
specified by this document.
When running tests, the event generation, among other data, shall be identified by a timestamp (specified in
Annex B), to detect possible delay between the time period of the detection process and the event generation.
7.2.4.3 Error handling and redundancy
Missed detections shall be logged and flagged for review.
7.3 Vehicle identification interfaces
In image-based EFC systems, a vehicle identification interface is identified between the ANPR camera and the
TCLS. ANPR cameras may be used to capture both front and rear licence plates of the vehicle. Cameras may
also capture pictures of the complete vehicle (front and rear), to support verification of the vehicle
identification by additional detected information, e.g. the make and the colour of the vehicle, and to compare
them to the officially registered information that is associated to the licence plate.
7.4 Vehicle classification interfaces
7.4.1 General
Vehicle classification interfaces are identified between the following physical objects:
— Vehicle classification camera and TCLS. The camera sends pictures of the vehicle to the TCLS and the
TCLS uses the pictures possibly supported by artificial intelligence (AI) to classify the vehicle (e.g.
determining its size and number of axles). This enables the toll charger (TC) to verify that the vehicle
detected data correspond to those officially registered.
— Vehicle classification radar or lidar and TCLS. The radar or lidar sends vehicle detected data to the
TCLS and the TCLS uses that data (possibly supported by AI) to classify the vehicle (e.g. determining its
size and number of axles). This enables the TC to verify that the vehicle detected data correspond to those
officially registered.
7.4.2 Event generation
Upon vehicle detection, the SUT shall generate an event containing:
— a timestamp,
— a lane identifier,
— the detection status (e.g. success, failure),
— the utilized detection method.
If specific characteristics are used for the classification method (e.g. number of axles) the information shall be
provided in the generated event.
If a national classification scheme exists, and it is different from the one used by the TCLS, both classifications
shall be provided in the generated event, each of them associated with an accuracy evaluation and the method
utilized (e.g. the national scheme can be derived by an algorithm applied to the TCLS classification).
7.4.3 Interface requirements
7.4.3.1 Classification mechanism
The classification mechanism shall be described by the SUT.
In case of an AI, support vector machine (SVM) or neural network detection process, the basic logic behind the
classification process shall also be declared and described.
NOTE If the classification process is based on national characteristics (e.g. specific markings fixed to the vehicle,
sequence of letters in the licence plate), the TCLS is generally not able to operate on foreign vehicles.
7.4.3.2 Data format and transmission
Classification events shall be transmitted in a specified format to the central processing unit in real-time, as a
single event or as a part of the picture metadata.
The test generated event shall be identified by a timestamp, to detect possible delays between the time period
of the classification process and the event generation. The format of the test generated events is specified in
Annex B.
7.4.3.3 Error handling and redundancy
Missed classification shall be logged and flagged for review.
In case of a classification score, an acceptability threshold shall be declared by the manufacturer, and all the
transits below the acceptability threshold shall be flagged for review.
8 Test purposes
8.1 General
Test purposes are specified in this document for the three capabilities of an image-based EFC system to detect
the passage of vehicles, to identify vehicles, and to classify vehicles.
These three capabilities are supposedly be tested independently, i.e. irrespective of the image-based EFC
system architecture.
The abbreviations in Table 1 will be used for naming the test purposes, see 8.2.
8.2 Naming
Test purposes are named in the following as XXYYYZZZ, where:
— [XX] indicates the process name, as per Table 1 ;;
— [YYY] indicates the variable name, as per Table 2 ;;
— [ZZZ] indicates the sequence number of the test purpose.
8.3 Format
Test purposes are specified in a table format, as shown by the proforma in Table 4, where the words in
boldface are entry names.
Table 4 — Test purpose proforma
Test Purpose name Test purpose description
Initial conditions Initial conditions
Repetitions Number of repetitions of the test purpose
Step sequence Step description
Sequence number Description of the step
Expected results Results expected. They may be expressed in terms of, e.g., number or percentage of true
positive results.
8.4 Laboratory tests
8.4.1 General
The following tests are intended to be performed in a closed laboratory with the characteristics and setups as
specified in 6.2. Laboratory tests are only aimed at identifying the vehicle by correctly reading its licence plate
([see Identification (of a licence plate)).)].
All laboratory tests shall be conducted with:
— camera aligned frontally to the licence plate (a limited lateral offset is allowed to cater for a lane width);
— installation height = manufacturer declared height or max laboratory height;
— licence plate distance configured to maintain the declared actual distance (AD).
8.4.2 Test setup and geometries
The following geometric parameters shall be declared by the manufacturer and used for testing (see Figure
2):
— H: maximum height of the camera,
— AD: actual distance between camera and licence plate,
— LO: lateral offset between the camera and the licence plate,
— GD: ground distance between the licence plate and the plane perpendicular to the licence plate where the
camera is positioned.
Figure 2 — Geometric parameters for laboratory tests
The same geometrical terms apply for the evaluation of:
— RSW (Reference Space Width at 0 lux),
— Identification depth, the maximum GD for which valid identification is verified,
— Limit positions used in static recognition tests. These are the points, in the coordinates (x,y,z) as shown
in Figure 2 ,, where the plate is positioned:
— The centre lane position at ground level (GD,0,0)
— The centre lane position at half of the maximum height of the camera (GD, 0, H/2)
— The centre lane position at maximum height of the camera (GD, 0, H)
— The right displacement from the centre lane at ground level (GD, LO, 0)
— The right displacement from the centre lane at half of the maximum height of the camera (GD, LO, H/2)
— The right displacement from the centre lane at maximum height of the camera (GD, LO, H)
— The left displacement from the centre lane at ground level (GD, -LO, 0)
— The left displacement from the centre lane at half of the maximum height of the camera (GD, -LO, H/2)
— The left displacement from the centre lane at maximum height of the camera (GD, LO, -H)
8.4.3 Static identification tests
Static identification tests, specified in Table 5 ,, Table 6 and Table 7 ,, are aimed at measuring the ability of
identifying static licence plates in different light conditions.
Table 5 — TP VILSI001: Identification of static licence plate in the laboratory in 0 lux illumination
conditions
VILSI001 Identification of vehicles in 0 lux light conditions
Initial conditions Position each licence plate at the nine limit points within the RS.
Consider three azimuth displacements for the licence plate: 0°, +αmax, -αmax.
Set illuminance to 0 lux.
Repetitions Repeat each step for each limit point and each azimuth angle for 100 times, for a total
of 2700 identifications
Step sequence Step description
1 Position the licence plate at one limit point at one azimuth displacement
2 Identify the licence plate and save the result
Expected results For each position and each azimuth displacement, the percentage of valid
identifications shall be better or equal to the expected value.
Table 6 — TP VILSI002: Identification of static licence plate in the laboratory in 10000 lux illumination
conditions
VILSI002 Identification of vehicles in 10000 lux light conditions
Initial conditions Position each licence plate at the nine limit points within the RS.
Consider three azimuth displacements for the licence plate: 0°, +αmax, -αmax.
Set illuminance to 10000 lux.
Repetitions Repeat each step for each limit point and each azimuth angle for 100 times, for a total
of 2700 identifications
Step sequence Step description
1 Position the licence plate at one limit point at one azimuth displacement .
2 Identify the licence plate and save the result
Expected results For each position and each azimuth displacement, the percentage of valid
identifications shall be better or equal to the expected value.
Table 7 — TP VILSI003: Identification of static licence plate in the laboratory in the presence of
shadows
VILSI003 Identification of vehicles in the presence of shadows
Initial conditions Position each licence plate at the nine limit points within the RS.
Consider three azimuth displacements for the licence plate: 0°, +αmax, -αmax.
Set two difference illuminances for two sectors of the licence plate, one to 10000 lux
and another one to 100 lux, in such a way to divide the licence plate into two
differently lighten areas. Prepare two possibilities: one where the licence plate is
differently shadowed vertically and one where the licence plate is differently
shadowed horizontally.
VILSI003 Identification of vehicles in the presence of shadows
Repetitions Repeat each step for each limit point, each azimuth angle and each for every mode of
shadowing for 100 times, for a total of 5400 identifications
Step sequence Step description
1 Position the licence plate at one limit point at one azimuth displacement with one
mode of shadowing.
2 Identify the licence plate and save the result
Expected results For each position, each azimuth displacement and each mode of shadowing, the
percentage of valid identifications shall be better or equal to the expected value.
8.4.4 Dynamic identification tests
Dynamic identification tests, specified in Table 8 and Table 9 ,, are aimed at measuring the ability of identifying
licence plates moving at different speeds with respect to the image-based EFC system camera.
Table 8 — TP VILDS001: Identification in the lab of a licence plate in motion at given speeds with
respect to the image-based EFC system camera
VILDS001 Identification of moving licence plates
Initial conditions Position the licence plate in the motion simulation system at the limit points within the
RS.
Set the motion simulation system to operate at two different speeds: 26Kmh and
70Kmh.
Set illuminance to 0 lux.
Repetitions Repeat each step for each limit point for each speed for 100 times, for a total of 1800
identifications
Step sequence Step description
1 Position the licence plate at one limit point and set the speed of the motion simulation
system.
2 Identify the licence plate and save the result
Expected results For each position and each speed, the percentage of valid identifications shall be better
or equal to the expected value.
Table 9 — TP VILDS002: Identification in the lab of a licence plate in motion at maximum speed with
respect to the image-based EFC system camera
VILDS002 Identification of moving licence plates
Initial conditions Position the licence plate in the motion simulation system at the limit points within the
RS.
Set the motion simulation system to operate at the maximum declared identification
speed.
Set illuminance to 0 lux.
Repetitions Repeat each step for each limit point, for a total of 900 identifications
Step sequence Step description
1 Position the licence plate at one limit point and set the speed of the motion simulation
system to the maximum declared identification value.
2 Identify the licence plate and save the result
VILDS002 Identification of moving licence plates
Expected results For each position, the percentage of valid identifications shall be better or equal to the
expected value.
8.5 On site passage detection
8.5.1 General
This group of tests is aimed at assessing the capability of the system to detect the passage of a vehicle through
the test area. All these tests are supposedly performed on site, whether in a close environment (test site) or in
an open environment (real live roads).
When multiple vehicles are involved in a test purpose, some of them may be substituted with projected images.
NOTE This practice assists in increasing the test subjects and reducing cost and complexity of the test.
8.5.2 Test setup
The test setup shall be as described in 6.4.
8.5.3 Detection of true positives
8.5.3.1 Detection by plate reading
Detection by licence plate reading test purpose, specified in Table 10, aims at measuring the ability of the IBES
of detecting passing vehicles by reading licence plates.
Table 10 — TP PDDET001: Detection of number of passing vehicles by licence plate reading
PDDER001 Detection of number of passing vehicles by plate reading
Initial conditions A vehicle representative of each category of vehicles liable for tolling according to the
local regulations and requirements is positioned near-by the test area, for example
(list is not exhaustive):
— 1 motorcycle
— 1 car
— 1 light truck (under X Ton.)
— 1 heavy or articulated truck (double license plate for truck and lorry)
Repetitions 10 repetitions. For each repetition, shuffle the order of passage of the objects.
Step sequence Step description
1 Let each vehicle pass through the system at the minimum speed allowed, with an
interval of X seconds between each transits
2 Let each vehicle pass through the system at the average speed allowed, with an
interval of X seconds between each transits
Let each vehicle pass through the system at the maximum speed allowed, with an
interval of X seconds between each transits
a) Number of vehicles detected as a percentage of actual vehicles passed.
Expected results
b) Images of the read licence plates
8.5.3.2 Detection independent of plate reading
Detection independent of licence plate reading test purpose, specified in Table 11, aims at measuring the
ability of the IBES of detecting passing vehicles without reading their licence plates.
Table 11 — TP PDDET002: Detection of number of passing vehicles independent of plate reading
PDDET002 Detection of number of passing vehicles independently of licence plate
reading
Initial conditions A vehicle representative of each category of vehicles liable for tolling according to the
local regulations and requirements is positioned near-by the test area, for example
(list is not exhaustive):
— 1 motorcycle
— 1 car
— 1 light truck (under X ton.)
— 1 heavy or articulated truck
Vehicles' licence plates are obscured or removed altogether.
Repetitions 10 repetitions. For each repetition, shuffle the order of passage of the objects.
Step sequence Step description
1 Let each vehicle pass through the system at the minimum speed allowed, with an
interval of X seconds between each transits
2 Let each vehicle pass through the system at the average speed allowed, with an
interval of X seconds between each transits
3 Let each vehicle pass through the system at the maximum speed allowed, with an
interval of X seconds between each transits
Expected results Number of vehicles detected as a percentage of actual vehicles passed.
8.5.4 Detection: false positives
False positives test purpose, specified in Table 12, aims at detecting vehicles intermixed with non-vehicle
passing objects. Non-vehicle objects are used to simulate false positives.
Table 12 — TP PDDFP001: Ability of not detecting not vehicle objects
PDDFP001 Ability of not detecting not vehicle objects
Initial conditions A vehicle representative of each category of vehicles liable for tolling according to the
local regulations and requirements is positioned near-by the test area, including at
least one not-vehicle object, for example (list is not exhaustive):
— 1 person
— 1 bicycle
— 1 motorcycle
— 1 car
— 1 light truck (under X Ton.)
Repetitions 10 repetitions. For each rep
...