JPEG Call for Information on Still Image Coding
WG1 is interested in the current state of the art for the coding of still images. ISO/IEC 10918 (JPEG-1) is by far the most widely used still image format despite being over twenty five years old. Several newer and more efficient still image coding systems have been standardized, some of which have found success in niche markets.
JPEG would like to collect information on these new codecs, both from WG1 and from other sources, to understand how these codecs operate and how they either are used or may be used commercially. While WG1 is interested in coding efficiency, coding efficiency is one of many important features for successful deployment of next generation still image coding tools. WG1 would like to use the information collected to determine opportunities for potential next generation still image codecs.
This investigation may lead to a new standardization effort within the framework of JPEG.
JPEG is interested in understanding the available still image coding technologies, their applications, requirements and use cases, their performance, feature sets and characteristics.
Contributions should provide an explanation of the basis of the compression algorithm and are asked to identify how the following features are, or might be, addressed by the contributed technology:
- Spatial accessibility (regions of interest)
- Resolution accessibility (access at different scales)
- Amenability to subjective optimization (pre- or post-compression)
- Ability to address high precision content, HDR content, very large format content, multi- and hyper spectral content, alpha channel, multi-layer, etc.
- Security and privacy features
- Compatibility to other existing image or video compression standards.
Additionally, contributors are asked to comment on the complexity of the scheme, in the following terms:
- Memory complexity
- Computational complexity (approximate number of operations depending on image size)
- Software complexity, including potential execution speed indicating the used platform and conditions
- Hardware complexity, to the extent that specific information may be available
- Asymmetry between encoding and decoding
3 Test Methods
In responding to this call for information, no specific test methodologies are imposed. Contributors should feel free to use any appropriate content, subjective assessment and objective metrics, either to assess compression efficiency, or to assess any of the other features in the list presented in the previous section. It is however encouraged if the content and subjective quality evaluation methodologies and objective quality metrics.
4 Features and Characteristics
Execution speed (software complexity)
Execution speed refers to a measure of time that takes to either encode or decode an image in a software implementation of the encoder or decoder.
Hardware complexity refers to any objective metric (e.g. number of gates, …) providing indications about the hardware complexity of an encoder or a decoder when implemented in specific or generic hardware architecture(s).
Memory requirements refer to any objective metric (e.g. amount of memory, type of memory, memory bandwidth, …) providing indications about memory requirements of an encoder or a decoder implemented in specific or generic hardware or software architecture(s).
Backward compatibility refers to possibility of decoding a bitstream generated by an encoder with of or more decoders compliant with existing JPEG standards.
Quality/Resolution scalability (regions of interest)
This refers to the possibility of partial decoding of a bitstream to produce an image with lower resolution (smaller size in pixels) or lower quality of either the entire image or a specific region of it (region of interest).
Amenability to subjective optimization (pre- / post-compression)
This refers to the possibility of optimization of either the encoding process directly, or by means of adequate pre- or post-processing that aim at improving the subjective quality of image either in the generated bitstream or in the decoder, or both.
Ability to address high precision content, HDR/WCG content, very large format content, multi- and hyper spectral content, alpha channel, multi-layer, etc.
This refers to ability to efficiently cope with content with enhanced characteristics such as High Dynamic Range, Wide Color Gamut, very large dimensions, multi- or hyper-spectral components, alpha channel(s), etc.
Security and privacy features
This refers to efficient embedding or security and privacy features in the encoding or decoding process, or in the bitstream.
|Call for Information||Feb 23 2015|
|Responses||Jun 22 2015|
|Report||Oct 07 2015|
Thomas Richter ()
Walt Husak ()
Touradj Ebrahimi ()