ISO/IEC 15444-3:2002/FDAM 3
(Amendment)Information technology — JPEG 2000 image coding system — Part 3: Motion JPEG 2000 — Amendment 3: Definition of compliance points and testing for Motion JPEG 2000
Information technology — JPEG 2000 image coding system — Part 3: Motion JPEG 2000 — Amendment 3: Definition of compliance points and testing for Motion JPEG 2000
Technologies de l'information — Système de codage d'image JPEG 2000 — Partie 3: Motion JPEG 2000 — Amendement 3: Définition des points et des essais de conformité pour Motion JPEG 2000
General Information
Relations
Standards Content (Sample)
FINAL ISO/IEC
AMENDMENT
DRAFT 15444-3
FDAM 3
ISO/IEC JTC 1
Information technology — JPEG 2000
Secretariat: JISC
image coding system —
Voting begins on:
2005-01-27
Part 3:
Motion JPEG 2000
Voting terminates on:
2005-03-27
AMENDMENT 3: Definition of compliance
points and testing for Motion JPEG 2000
Technologies de l'information — Système de codage d'image
JPEG 2000 —
Partie 3: Motion JPEG 2000
AMENDEMENT 3: Définition des points et des essais de conformité
pour Motion JPEG 2000
Please see the administrative notes on page iii
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 SUPPORT-
ING DOCUMENTATION.
IN ADDITION TO THEIR EVALUATION AS
Reference number
BEING ACCEPTABLE FOR INDUSTRIAL, TECHNO-
ISO/IEC 15444-3:2002/FDAM 3:2005(E)
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. ISO/IEC 2005
---------------------- Page: 1 ----------------------
ISO/IEC 15444-3:2002/FDAM 3:2005(E)
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.
Copyright notice
This ISO document is a Draft International Standard and is copyright-protected by ISO. Except as permitted
under the applicable laws of the user's country, neither this ISO draft nor any extract from it may be
reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic,
photocopying, recording or otherwise, without prior written permission being secured.
Requests for permission to reproduce should be addressed to 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
Reproduction may be subject to royalty payments or a licensing agreement.
Violators may be prosecuted.
ii © ISO/IEC 2005 – All rights reserved
---------------------- Page: 2 ----------------------
ISO/IEC 15444-3:2002/FDAM 3:2005(E)
In accordance with the provisions of Council Resolution 21/1986, this document is circulated in the
English language only.
© ISO/IEC 2005 – All rights reserved iii
---------------------- Page: 3 ----------------------
ISO/IEC 15444-3:2002/FDAM 3:2005(E)
Contents Page
Foreword .v
Introduction.v
8 Definition of compliance points.1
8.1 H, W, C: Image size guarantees .1
8.2 N : Code-block parsing guarantee .2
cb
8.3 N : Component parsing guarantee .2
comp
8.4 L : Coded data buffering guarantee.2
body
8.5 M: Decoded Bit-plane guarantee .3
8.6 P: 9-7I Precision guarantee .3
8.7 B: 5-3R Precision guarantee .3
8.8 T : Transform level guarantee.3
L
8.9 L: Layer guarantee .4
8.10 Progressions.4
8.11 Tiles.4
8.12 Tile-parts.4
8.13 Precincts.4
8.14 Frame-rate and bit-rate .4
8.15 Profile: codestream guarantee.4
9 Compliance point definitions.4
10 Definition of test methods .5
11 Executable test suite (ETS) .6
11.1 Test sequences.6
11.2 Cpoint-3.7
11.3 Cpoint-2.8
11.4 Cpoint-1.9
11.5 Cpoint-0.10
Annex E (informative) Guidelines for Implementing Motion JPEG 2000.11
Annex F (informative) Guide to JPEG 2000.14
Annex G (informative) Reference components file format .16
iv © ISO/IEC 2005 – All rights reserved
---------------------- Page: 4 ----------------------
ISO/IEC 15444-3:2002/FDAM 3:2005(E)
Foreword
ISO (the International Organization for Standardization) and IEC (the International Electrotechnical
Commission) form the specialized system for worldwide standardization. National bodies that are members of
ISO or IEC participate in the development of International Standards through technical committees
established by the respective organization to deal with particular fields of technical activity. ISO and IEC
technical committees collaborate in fields of mutual interest. Other international organizations, governmental
and non-governmental, in liaison with ISO and IEC, also take part in the work. In the field of information
technology, ISO and IEC have established a joint technical committee, ISO/IEC JTC 1.
International Standards are drafted in accordance with the rules given in the ISO/IEC Directives, Part 2.
The main task of the joint technical committee is to prepare International Standards. Draft International
Standards adopted by the joint technical committee are circulated to national bodies for voting. Publication as
an International Standard requires approval by at least 75 % of the national 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 and IEC shall not be held responsible for identifying any or all such patent rights.
Amendment 3 to ISO/IEC 15444-3:2002 was prepared by Joint Technical Committee ISO/IEC JTC 1,
Information technology, Subcommittee SC 29, Coding of audio, picture, multimedia and hypermedia
information.
© ISO/IEC 2005 – All rights reserved v
---------------------- Page: 5 ----------------------
ISO/IEC 15444-3:2002/FDAM 3:2005(E)
Introduction
ITU-T Rec.T.800 | ISO/IEC 15444-1 is a specification that describes an image compression system that
allows great flexibility, not only for the compression of images but also for access into the codestream. ITU-T
Rec.T.802 | ISO/IEC 15444-3 specifies the use of the wavelet-based JPEG 2000 codec for the coding and
display of timed sequences of images. The Motion JPEG 2000 file format (MJ2) is designed to contain one or
more motion sequences of JPEG 2000 images, with their timing, and also optional audio annotations, all
composed into an overall presentation. ITU-T Rec.T.803 | ISO/IEC 15444-4 provides the framework, concepts,
and methodology for testing and the criteria to be achieved to claim compliance to JPEG 2000 standard, i.e. to
still images only.
This document makes use of the latter framework of conformance testing to apply it to the mentioned motion
sequences. The objective of standardization in this field is to promote interoperability between MJ2 encoders
and decoders and to test these systems for compliance to these specifications. Compliance testing is the
testing of a candidate product for the existence of specific characteristics required by a standard. It involves
testing the capabilities of an implementation against both the compliance requirements in the relevant
standard and the statement of the implementation’s capability.
With this document the framework, concepts, methodology for testing and the criteria to be achieved to claim
compliance to ITU-T Recommendation T.802 | ISO/IEC 15444-3 are specified. Whilst extending the
specifications of ITU-T Rec.T.803 | ISO/IEC 15444-4 to timed sequences of images, the mind was on widely
used applications. This document describes compliance points and testing procedures for Motion JPEG 2000
decoders only.
vi © ISO/IEC 2005 – All rights reserved
---------------------- Page: 6 ----------------------
ISO/IEC 15444-3:2002/FDAM 3:2005(E)
Information technology — JPEG 2000 image coding system —
Part 3:
Motion JPEG 2000
AMENDMENT 3: Definition of compliance points and testing for
Motion JPEG 2000
Add the following reference to clause 2:
ITU-T Rec.T.803 | ISO/IEC 15444-4: Information technology — JPEG 2000 image coding system: Conformance
testing
Add the following clauses after clause 7:
8 Definition of compliance points
This section describes a number of compliance points (Cpoints) for ITU-T Rec.T.802 | ISO/IEC 15444-3. The
points and parameters are described to provide assistance in designing a compliant decoder. Actual
compliance is determined by the test methods in clause 10 and the codestreams, reference images, and
tolerances in clause 11. The definitions of compliance points in this section are useful for the design of an
encoder. The parameters may correspond to particular parts of an implementation.
Because of resource limitations, implementations of Motion JPEG 2000 sometimes will not be able to decode
a codestream in its entirety. This section defines various parameters for which a specific implementation might
be limited. A set of values for every parameter defines a compliance point. Thus an implementation of a
particular Cpoint must guarantee resources as defined in all the parameters.
8.1 H, W, C: Image size guarantees
Decoders may be limited in the size of the output image that they are capable of producing, due to physical
display characteristics or memory limitations. H, W, and C are respectively the largest height, width, and
number of components that are required to be decoded for a decoder in the compliance point. Codestreams
containing more samples than the H, W, and C for a Cpoint shall still be decoded, provided they contain a
resolution equal to or less than HxW. Compliance for these codestreams is based on the ability to decode at
the largest size smaller than or equal to that specified by the decoder’s Cpoint, while preserving aspect ratio.
The requested image size is defined by the height and width fields in the applicable ‘VisualSampleEntry’ from
the MJ2 file.
Equation 8.1 and Equation 8.2 express these restrictions. The maximum that satisfies both conditions is
r ≤ T
L
the number of levels that must be decoded. T is defined in subclause 8.8. The variables w and h denote
L r r
requested with and height from the VisualSampleEntry segment defined in subclause 6.1. The term
min(N (i))
denotes the minimum number of decomposition levels declared in any COD or COC marker
L
∀i
segments whether used in main header or in tile-part headers as defined in Annex A of ITU-T Rec.T.800 |
ISO/IEC 15444-1.
© ISO/IEC 2005 – All rights reserved 1
---------------------- Page: 7 ----------------------
ISO/IEC 15444-3:2002/FDAM 3:2005(E)
If a non-negative r does not exist to satisfy both conditions for any tile or for the whole image, then no decoder
obligation exists. A decoder claiming compliance at some Cpoint with image dimensions H x W and number of
components C, must also be capable of decoding any sequences with width less than or equal to W, height
less than or equal to H, and number of components less than or equal to C. For each Cpoint, the minimum
values for H, W, and C are specified in table AMD3-1.
w
r
≤ W 8.1
min( N (i))−r
L
∀i
2
h
r
8.2
≤ H
min( N (i))−r
L
∀i
2
8.2 N : Code-block parsing guarantee
cb
Decoders need not decode compressed bits that cannot be recovered from the codestream due to excessive
parser memory being required. An upper bound for the parser state memory required to reach a point x in the
codestream may be determined from the total number of code-blocks for which state information must be kept,
the total number of precincts for which a packet has been encountered, and the total number of components
of the codestream.
At position x in the codestream, N (x) is defined as the total number of code-blocks in every precinct where
cb
the first header byte of at least one received packet for the precinct lies outside the range 0x80 to 0x8F.
Decoders are permitted to stop parsing the codestream at the point, x, once N (x) > N , where N is defined
cb cb cb
for each compliance point. Decoders are permitted to stop parsing the codestream once packet headers with
th
more than N code-blocks have been encountered. Code-blocks in packets prior to the packet with the N
cb cb
code-block shall be decoded up to the limits of other parameters in the compliance point.
NOTE — Packets headers with the first bit set to 0 are defined as empty. The above definition adds all the code-blocks
associated with such precincts to N for these empty packets because a decoder requires more memory for these
cb
packets than for packets starting in the listed range.
8.3 N : Component parsing guarantee
comp
Decoders could be required to buffer information about each component for many thousands of components
just to parse a codestream. To limit the required memory, decoders are permitted to stop parsing the
codestream at a point, x, once the following condition is reached:
C (x) > N
max comp
where C (x) is defined as the largest component index for which a packet has been encountered up to point
max
x regardless of the emptiness or the relevance of the packet.
Code-blocks in packets prior to the above stop condition shall be decoded up to the limits of other parameters
in the compliance point.
8.4 L : Coded data buffering guarantee
body
The parser state memory described in subclause 8.2 is required to parse packets regardless of whether their
code-blocks are relevant to the dimensions and number of components for which compliance is being claimed.
For those code-blocks that are relevant, the implementation is required to store the recovered packet bytes.
These are the code bytes that are processed by the block decoder (Annexes C and D, ITU-T Rec.T.800 |
ISO/IEC 15444-1).
After a given number of decoded codestream bytes, x, the quantity L (x) is defined as the total number of
body
packet bytes that have been encountered so far in packets whose precincts are relevant to the dimensions
2 © ISO/IEC 2005 – All rights reserved
---------------------- Page: 8 ----------------------
ISO/IEC 15444-3:2002/FDAM 3:2005(E)
and components for which compliance is being claimed. Although some implementations may be able to
decode some of these packet bytes incrementally, L represents an upper bound on the number of packet
body
bytes that must be stored by the decoder prior to decoding. If the number of relevant packet bytes exceeds
L , then the Implementation Under Test (IUT) is allowed to stop reading the codestream and to decode the
body
code-blocks obtained up to the limits of other parameters in the compliance point.
8.5 M: Decoded Bit-plane guarantee
The decoder shall decode all of the packet bytes recovered by the parser in accordance with the requirements
described above. This obligation is limited to the most significant M bit-planes of each code-block. Specifically,
the block decoder must correctly decode the first coding passes, if available, of any relevant
3(M − P ) − 2
B
code-block, b, where P is the number of zero-valued most significant bit-planes signaled in the relevant
b
packet header as described in Annex B of ITU-T Rec.T.800 | ISO/IEC 15444-1. The decoder is free to decode
any number of additional coding passes for any code-block. Codestreams with large values for the number of
guard bits will have a larger number of zero-valued most significant bit planes, and thus a decoder of any
given Cpoint will decode fewer useful bit-planes. Likewise, codestreams with large values for the shift in the
RGN marker segment may have fewer bit-planes decoded.
8.6 P: 9-7I Precision guarantee
Codestreams that make use of the irreversible 9-7 discrete wavelet transform will require dequantization, the
9-7 inverse discrete wavelet transform, and potentially the inverse irreversible component transform (ICT).
The precision values for the wavelet transform are chosen to allow high quality imagery at various bit-depths,
e.g. 8, 12, or 16 bits per sample. However, for Cpoint-0, the accuracy of the 9-7I filter required is set such that
it is possible to be compliant by decoding and inverse quantizing and performing a 5-3I (irreversible 5-3)
inverse wavelet transform. This allows lower cost decoders to be used for the lowest compliance point only.
For higher compliance points, using the 5-3 filter in place of the 9-7 filter will not be sufficient to pass the
compliance tests.
Using the 5-3 inverse wavelet transform to decode imagery compressed with the 9-7 wavelet introduces signal
dependent noise. For example errors are highest around edges in the imagery. Because induced errors are
signal dependent, there is no “precision” specified for the implementation of the wavelet transform for Cpoint-0.
Instead, the bounds on accuracy of the 9-7 transform have been set for each Cpoint-0 reference image to
allow an implementation to use the 5-3I inverse wavelet filter. Using the 5-3I inverse wavelet transform instead
of a 9-7I filter does not relieve a decoder of the requirement to perform inverse quantization.
For compliance points other than Cpoint-0, the precision guarantee in Table 1 refers to the implementation’s
minimum word size that will achieve the target MSE values for the test streams.
To facilitate end-to-end testing for compliance, dequantization may be performed using mid-point rounding.
That is, the value of r in Equation G.6 of ITU-T Rec.T.800 | ISO/IEC 15444-1 can be r = 1/ 2 . Implementations
under test may provide the option of using different values for the reconstruction parameter, r; however, if the
value r = 1/ 2 is supported and employed for compliance testing this will typically increase the ease of passing.
8.7 B: 5-3R Precision guarantee
A decoder is expected to implement the reversible 5-3R IDWT exactly, for component bit-depths of B
bits/sample or less, as specified in the SIZ marker segment (see Annex A of ITU-T Rec.T.800 |
ISO/IEC 15444-1). If a codestream employs the reversible component transform (RCT) and the IUT claims
compliance at 3 or more components, it must be able to perform both the 5-3R IDWT and the inverse RCT
exactly for bit-depths of B bits/sample or less.
8.8 T : Transform level guarantee
L
For each Cpoint, a decoder is expected to be able to synthesize a minimum number of levels of the IDWT, T .
L
For codestreams that contains more than T decomposition levels, the decoded image from a compliant
L
decoder in a given Cpoint may include only the top resolution levels.
© ISO/IEC 2005 – All rights reserved 3
---------------------- Page: 9 ----------------------
ISO/IEC 15444-3:2002/FDAM 3:2005(E)
8.9 L: Layer guarantee
For each Cpoint, a decoder is expected to decode a minimum number of layers, L, in a codestream. For
codestreams that contains more than L layers, the decoded image from a compliant decode in a given Cpoint
may include only the top L layers. This relieves compliant decoders from the burden of decoding inefficient
codestreams with an excessive number of layers.
8.10 Progressions
For all Cpoints, a decoder is expected to decode all possible progressions as specified in the COD marker
segment. If a POC marker segment is used in a codestream, Cpoint-0 to Cpoint-3 decoders shall decode
packets associated with the first progression order specified in the POC marker segment for that tile.
Additional packets in the tile may be skipped.
8.11 Tiles
If an image is divided into tiles the following restrictions apply to tile dimensions:
i i
128 ≤ XTsiz / min(XRsiz, YRsiz) ≤ 1024 8.3
XTsiz = YTsiz 8.4
8.12 Tile-parts
Codestreams may contain multiple tile-parts for each tile. Profile-0 codestreams require all initial tile-parts to
appear in spatial order in the codestream before other tile parts. Cpoint-0 to Cpoint-3 decoders may ignore
tile-parts beyond the first even if N or L has not been reached.
cb body
8.13 Precincts
Tiles may contain several precincts. Cpoint-0 decoders need only decode the first precinct in each subband of
each tile.
8.14 Frame-rate and bit-rate
Frame-rate: A compliant real-time decoder must report the lowest frame rate that it can sustain when
decoding all frames, as well as the number of skipped frames (fields) when it achieves real-time.
Bit-rate: A compliant real-time decoder must also report the highest bit-rate which can always be fully decoded
in real-time (bit-rate guarantee).
8.15 Profile: codestream guarantee
Profiles provide limits on the codestream syntax parameters. Two profiles are defined in ITU-T Rec.T.802 |
ISO/IEC 15444-3, labeled ‘unrestricted’ and ‘simple’. Conformance testing of the rich feature set of
unrestricted codestreams is not targeted in this document. Thus, for all Cpoints compliant decoders need only
to handle motion representations in MJ2 simple profile, indicated by the brand ‘mj2s’ in top-level file-type box.
NOTE — Conforming to simple profile means that restrictions of Profile-0 defined in Annex A.10 of ITU-T
Rec.T.800 | ISO/IEC 15444-1 apply to embedded codestreams, with the exception of tile dimensions where
YTsiz = XTsiz = 128 is replaced by definitions of subclause 8.11
9 Compliance point definitions
Table AMD3-1 defines four compliance points in terms of the parameters.
4 © ISO/IEC 2005 – All rights reserved
---------------------- Page: 10 ----------------------
ISO/IEC 15444-3:2002/FDAM 3:2005(E)
Table AMD3-1 — Definitions of compliance points (Cpoint) for Part-3
Parameter Cpoint-0 Cpoint-1 Cpoint-2 Cpoint-3
WxH(Size) 360x288 720x576 1920x1080 4096x3112
C(Components) 3 3 4 4
N 399 1371 8428 50656
cb
N 4 4 4 4
comp
17 20 23 26
L 2 bytes 2 bytes 2 bytes 2 bytes
body
M 11 13 15 19
P Low enough to 16 bit fixed point 16-bit fixed point 20 bit fixed point
allow 5x3 I implementation implementation implementation
decoding of 9x7 I
data
B 8 10 12 16
T 3 4 5 5
L
L 15 15 15 15
Progressions For all Cpoints, a decoder is expected to decode all possible progressions as specified in
the COD marker segment. If a POC marker segment is used in a codestream, a Cpoint-0
decoder shall decode packets associated with the first progression order specified in the
POC marker segment for that tile. Additional packets in the tile may be skipped. For all
other Cpoints, packets may be skipped only due to other limitations (e.g. N and L )
cb body
and there is no explicit limitation on the number of progression order changes that may
occur.
Tiles Single tile image Single tile image or Single tile image or Single tile image or
or square tiles square tiles with square tiles with square tiles with
with dimensions dimensions ranging dimensions ranging dimensions ranging
ranging from 128 from 128 to 1024 from 128 to 1024 from 128 to 1024
to 1024
Tile-parts Decode only first Decode only first tile- Decode only first tile- Decode only first tile-
tile-part per tile part per tile part per tile part per tile
Precincts Decode first Decode all precincts Decode all precincts Decode all precincts
precinct per sub-
band
File format MJ2 simple MJ2 simple profile MJ2 simple profile MJ2 simple profile
profile
10 Definition of test methods
Compliance testing procedures apply as defined in Annex B of ITU-T Rec.T.803 | ISO/IEC 15444-4, with
following extensions:
1. A particular executable test suite (ETS) defines the test codestreams (TCS), output images and error
tolerances. This is done in clause 11 for the four defined compliance points by taking specified frames
from MJ2 sequences. Implementations under test (IUT) must therefore be able to output decoded
visual samples in a format (see Annex G) that they can be compared as defined in ITU-T Rec.T.803 |
ISO/IEC 15444-4. Any visual composition transformations do not apply to this test method.
2. In addition to the testing of single visual samples IUT will be evaluated by using MJ2 reference
sequences defined in clause 11. The test procedure is:
© ISO/IEC 2005 – All rights reserved 5
---------------------- Page: 11 ----------------------
ISO/IEC 15444-3:2002/FDAM 3:2005(E)
• To decode all frames (fields) of a sequence as defined in subclause 8.14. Therefore it must report,
that no frames (fields) were skipped.
• To decode in correct frame order and correct field order. Therefore it must report, that the correct
frame (field) order has been respected.
• To report the lowest frame rate it can sustain as defined in subclause 8.14
• To report the highest bit-rate which can always be fully decoded as defined in subclause 8.14
Table AMD3-2 lists the obligations for an IUT evaluated to be compliant to a certain Cpoint.
Table AMD3-2 — Obligations for IUTs
Cpoint-0 TCS Cpoint-1 TCS Cpoint-2 TCS Cpoint-3 TCS
Cpoint-0 Must decode Must decode up to Must decode up to Must decode up to
IUT everything compliance point compliance point compliance point
parameters defined parameters defined parameters defined
in Cpoint-0 in Cpoint-0 in Cpoint-0
Cpoint-1 Must decode Must decode Must decode up to Must decode up to
IUT everything everything compliance point compliance point
parameters defined parameters defined
in Cpoint-1 in Cpoint-1
Cpoint-2 Must decode Must decode Must decode Must decode up to
IUT everything everything everything compliance point
parameters defined
in Cpoint-2
Cpoint-3 Must decode Must decode Must decode Must decode
IUT everything everything everything everything
11 Executable test suite (ETS)
This section defines four ETSs for the compliance points (Cpoints) defined in clause 9. Additional ETSs may
be made available after the publication of this standard, see http://www.jpeg.org/software for the latest set of
ETSs.
Each ETS consists of motion sequence and single image codestreams, reference decoded images, and
tolerance values for MSE and peak error.
11.1 Test sequences
The following compressed test sequences listed in Tables AMD3-3 to AMD3-5 are attached to ITU-T
Rec.T.802 | ISO/IEC 15444-3.
Table AMD3-3 — List of test sequences (Cpoint-3)
Sequence name Resolution Video Type Number of Frames Format Bit depth
cp31.mj2 4096x3112 24P 80 RGB 4:4:4 16
6 © ISO/IEC 2005 – All rights reserved
---------------------- Page: 12 ----------------------
ISO/IEC 15444-3:2002/FDAM 3:2005
...
Questions, Comments and Discussion
Ask us and Technical Secretary will try to provide an answer. You can facilitate discussion about the standard in here.