During the development of digital audio and video compression coding standards, Joint Photographic Experts Group, JPEG, Dynamic Image Experts Group, MPEG, and Video Coding Expertgroup, Vceg) It is a crucial driving role. What are the criteria for digital audio and video compression coding?
1, MJPEG and MJPEG2000 series
JPEG is a working group operating within the International Organization for Standardization, ISO and the International Telephone Telegraph Advisory Committee (Consultation Committee of the International Telephone and Telegraph, CCIT). In terms of video compression, JPEG has previously prepared MJPEG and MJPEG2000. MJPEG is a dynamic image compression technology developed in the JPEG foundation, which is only compressed for a frame separately. It is basically regardless of the change in different frames in the video stream. Use this technology to get a high-definition video image, and you can flexibly set the number of video definitions and compressed frames per channel. Its compressed screen can also be arbitrarily spliced. MJPEG video compression in the same format is different from inter-frame compression because the compression bit rate is relatively low, so encoding and decoding are relatively easy, and there is no need for too much calculation capability, which makes it easy to edit MJPEG. Because of this, some mobile devices, such as digital cameras using MJPEG to make short-slice encodings. However, its defects are also very obvious: 1. If you lose frame, the real-time nature is poor, it is difficult to complete real-time compression under the premise of ensuring that each path must be high, its second, low compression efficiency, large storage space.
MJPEG 2000 is a third part of the JPEG 2000 standard, which is a compression criterion for coding moving images on the basis of the first part of the standard. MJPEG 2000 is a standard for image sequences that support non-destructive and lossless compression in a separate codec. It allows one or more JPEG 2000 compressed image sequences and sound, metadata, stored as the file format of MJ2.
2.mpeg series
MPEG-1 is the first video and audio loss compression standard set by MPEG, and the earliest introduction and application of MPEG technology in the market, its original main goal is to record images on the CD CD, and later applied in VCD CD middle. At the end of 1992, MPEG-1 was officially approved to become an international standard. MPEG-1 compresses the standard resolution of Standard Image Format, SIF) (for NTSC system is 352x240; for PAL system is 352x288), the transmission rate is 1.5Mb / s, and 30 frames are played per second. With CD sound quality, the quality level is basically equivalent to the home recording system (VIDEO HOME SYSTEM, VHS). The MPEG-1 coding rate is up to 4 to 5Mb / s, but as the rate is increased, the quality of the image after decoding has decreased.
MPEG-2 is set in 1994, its design goal is to provide an image of a high-level industrial standard and a higher transmission rate. It is a coding standard for 3 to 10MB / s motion images and its sounds of HDTV and DVDs. The main features of MPEG-2 technologies include: simultaneous supporting interlaced inputs and progressive scanning inputs; providing a wide range of variable compression ratios to accommodate different picture quality, storage capacity, and bandwidth requirements. MPEG-2 solves all types of applications into different grades, grade, and level concepts, based on the complexity of video coding technology, to solve the exchangeability and internationality of the PROFILE and Level (Level), grade and level. MPEG-2 adds a graded coding property that allows different quality grades or different time-space resolutions from a coded data stream.
MPEG-4 officially became an international standard in early 2000. The design objective of MPEG-4 is to provide multimedia communication at low bit rates. MPEG-4 is more suitable for interactive AVS services and remote monitoring compared to previous criteria, more focusing on interactive and flexibility of multimedia systems. The compression method of MPEG-4 can select a suitable algorithm for system crop according to application requirements.
The purpose of MPEG-7 design is to solve contradictions between massive images and sound information and fast retrieval. MPEG-7 is known as multimedia content description interface, its goal is to generate a standard that describes multimedia content data, meets the needs of real-time, non-real-time, and push-pull applications. MPEG-7 extends the dedicated solutions and related functions of existing identification content, including more multimedia data types.
The goal of MPEG-21 design is to organically integrate different protocols, standards, and technologies, and develop new standards. MPEG-21 is committed to transferring and using defining a standardized, interoperable and highly automated open frame. The basic framework of MPEG-21 includes the identification and description of digital items, content representation, digital item description, content management and use, intellectual property management and protection, terminal, network, and event reporting.
3.H.26X series
H.261 is a video coding standard set by VCEG, belonging to the video codec section. H.261 is the first practical digital video coding standard. The purpose of its design is to transmit quality reliable video signals on a comprehensive business digital network with a bandwidth of 64kb / s. H.261 uses a hybrid encoding frame. H.261 is only strictly restrictive regulations related to compatibility related symbol grammar, symbol multiplexing, and decoding process, etc., and some of the restrictions on the recovery image quality indicators but does not affect compatibility. Provisions, providing developers, vendors, and users with a large application space.
H.263 is a standard draft of VC ++ EG, designed for low code streamline letters. It provides four optional coding algorithms: unlimited motion vector mode, advanced prediction mode, PB frame mode, and semantic arithmetic prediction mode. H.263 + is the second version of H.263. H.263 + provides 12 new negotiable modes and other features such as advanced intra coding, decking effect filtering, reference frame selection, SNR / time domain / null grade, etc., further improves compression encoding performance. H.263 + Allows the use of more source formats, there are also multiple options for image clock frequencies, broaden the scope of application; another - important improvements are scalability, which allows multiple settings rates and multi-resolution, enhanced video information Transport in the volatile, easy-to-package heterogeneous network environment, and also allows many additional information to be added within the stream, greatly facilitating the user's operation. H.263C ++ is the third version of H.263, H.263C ++ has added a lot of functions in the anti-indication of the video stream, increasing the reversible VLC encoding and the data classification-based anti-misdisper, and expands the reference frame selection Mode; enhanced the robustness of the anti-mistake system.
The H.26L design is to achieve low-bit rate, real-time, and low-delayed video coding for a variety of image sources. H.26L is the prototype of H.264. H.264 is also a tenth in MPEG-4. H.264 As a new generation of video compression algorithms, it has absorbed the advantages of various coding schemes, especially MPEG-2 and H.263C ++, and has made a lot of improvements in grammatical structures, encoding prediction algorithms, data transform output methods, etc. Performance has been greatly improved. Its codec process mainly includes 5 parts: inter-frame and intra prediction, transform, and reverse exchange, quantization, and reverse quantization, loop filtering, entropy encoding. H.264 has the characteristics of low code stream, high quality image, strong fault tolerance, strong network adaptability.
As a new generation of video coding standards, HEVC (H.265) is still a hybrid encoding framework for "predictive increase". However, relative to H.264, H.265 has a revolutionary change in many ways.
Flexible coding structure
In H.265, the size of the macroblock is extended from H.264 to 64x64 to facilitate compression of high resolution video format. At the same time, a more flexible encoding structure is used to increase coding efficiency, including coding units, and predicting the unit transform unit.
Flexible block structure - RQT
RQT (Residual Quad-Tree Transform) is an adaptive transformation technology that extends and extension of ABT (Adaptive Block-Size Transform) technology in H.264 / AVC. For interframe encoding, it allows the size of the transform block to adaptive adjustments according to the size of the motion compensation block; for intra coding, it allows the size of the transform block to adaptively adaptively in accordance with the characteristics of the intra prediction residual. adjust.
Big block transformations can provide better energy concentrations with respect to small blocks, and can save more image details after quantification, but on the other hand, there will be more ringing after quantification. effect. Therefore, the transform block size is adaptively selected depending on the characteristics of the current block signal.
Sampling point adaptive offset
SAO (SAMPLE Adaptive Offset) After the code loop, after DEBLOCK, by adding an offset to the reconstructed image, adding an offset to each type of image pixel value, reducing the purpose of reducing distortion, thereby increasing the compression ratio, reducing code flow. After SAO, the average can reduce the code stream of 2% to 6%, while the performance consumption of the encoder and decoder increases by about 2%.
Adaptive loop filtering
ALF (Adaptive Loop Filter) within the coded loop, after the DEBLOCK and SAO, used to restore the reconstructed image to achieve a wifter difference (MSE) between the reconstruction image and the original image. The coefficient of ALF is to apply ALF in frame-level computing and transmission, or may perform ALF based on blocks or quadtree, and if it is based on part of the ALF, it is necessary to pass indicator area information. Additional information.
Parallelization design
The current chip architecture has gradually developed from single nuclear performance, so in order to adapt to a very high degree of chip implementation, HEVC / H265 introduces many parallel operations, overcomes the defect of H.264.
LW, read full story
Our other product:
