Full HD video encoding means for high-definition video processing technology. Technically, the overall physical display full HD resolution to achieve 1920 × 1080P, that is, the horizontal resolution to 1920 pixels vertical resolution to achieve the 1080 scan lines. Meet the standards of full HD LCD TV panel in its label or as close to the TV will be part of the design, with
video image data related principles have a strong correlation, that a large number of redundant information. Redundant information which can be divided into spatial and temporal redundant information redundant information. Data compression technology is to remove redundant information (the correlation between the removal of data), image data compression techniques include intraframe compression, frame image data compression and entropy coding compression techniques.
redundant information to the time domain can be removed using the inter-frame encoding redundant information in time domain, which includes the following three parts:
-
motion compensation, motion compensation is through the local image to the previous forecast, the current compensation for the local image, which is to reduce redundant information frame sequence effective way.
-
movement that different regions need to use a different image to describe the motion information of motion vectors. Motion vector is compressed by entropy coding.
-
motion estimation motion estimation in video sequences taken from the set of motion information technology.
Note: use common compression standard block-based motion estimation and motion compensation
to the main use of spatial redundancy and entropy coding interframe coding techniques:
- transform image coding and intra prediction
differential signal has a high spatial redundancy. Transform coded signal into the airspace of another orthogonal vector space, so that correlation decreased, reducing data redundancy.
-
after quantization coding transform coding, the resulting number of transform coefficients, these coefficients are quantized, so that the output of the encoder to achieve a certain bit rate. This process leads to reduced accuracy.
- entropy coding
entropy coding is lossless coding. It transform, the quantized coefficients and motion information obtained, for further compression.
Edit this paragraph the basic framework of video coding standard H.261
H.261 for ISDN design, the main design for real-time encoding and decoding, compression and decompression of the signal delay of less than 150ms, Rate px64kbps (p = 1 ~ 30).
H.261 standard is mainly used for inter prediction of motion compensation, DCT transform, adaptive quantization, entropy coding compression technology. Only I frames and P frames, no B frames, motion estimation is only accurate to pixel level. Supports two image scanning formats: QCIF and CIF.
H.263
H.263 standard is even low bit rate image coding international standard, which on the one hand to H.261-based, hybrid coding as the core, the basic block diagram and H.261 is very similar to the original data and streams similar organizations; the other hand, H.263 has absorbed a number of international standards, MPEG and other valid and reasonable part, such as: half pixel accuracy motion estimation,
red wing shoes stores, PB frame forecast, it is better than H.261.
H.263 bit rate used is less than 64Kb / s, and the transfer time to a fixed bit rate (VBR). H.263 supports multiple resolutions: SQCIF (128x96), QCIF, CIF, 4CIF, 16CIF. H.261 and H.263
with relevant international standards
with relevant international standards H.261
H.320: narrow-band visual telephone systems and terminal equipment ;
H.221: audio-visual telecommunication services in the 64 ~ 1 920Kb / s channel frame structure;
H.230: Frame synchronous audio-visual system control and indication signals;
H.242: Use until the 2Mb / s digital channels of audio-visual terminal system.
with relevant international standards H.263
H.324: low bit rate multimedia communications terminal equipment;
H.223: low bit rate multimedia communication composite agreement;
H.245: Control Protocol for Multimedia Communication;
G.723.1.1: transmission rate of 5.3Kb / s and 6.3Kb / s speech encoder.
JPEG
International Organization for Standardization was established in 1986 JPEG (Joint Photographic Expert Group) Joint Photographic Experts Group, focuses on development of continuous-tone, multi-level gray scale, still image digital image compression standard. Commonly used discrete cosine transform (DCT) coding method is the core of JPEG algorithm.
MPEG-1 / 2
MPEG-1 standard for digital storage body coding of moving pictures and associated audio, the digital rate of 1.5Mb / s. MPEG-1 and H.261 video block diagram similar.
MPEG-1 video compression technology features: 1. random access; 2. Fast Forward / Reverse search; 3. reverse replay; 4. audio synchronization; 5. fault tolerance; 6. encoder / decoder delay. MPEG-1 video compression strategy: To improve the compression ratio, intra / inter-frame image data compression technique must be used. Intra-frame compression algorithm similar to JPEG compression algorithm, DCT-based transform coding techniques to reduce spatial redundancy. Interframe compression algorithm, the use of forecasting and interpolation. Forecast error can be handled through the DCT transform coding, and further compression. Interframe coding techniques can reduce the time axis direction of the redundant information.
MPEG-2 is known as the
MPEG-4
MPEG-4 standard MPEG-2 is not a substitute for, it focuses on different applications. MPEG-4 formulation designed mainly for video conferencing, video phone ultra-low bit-rate reduction (less than 64Kb / s) needs. In the development process, MPEG deeply impressed by people's organizations, media information, especially for the needs of video information from the playback type to content-based access, retrieval and operation.
MPEG-4 and the previously mentioned JPEG, MPEG-1 / 2 are very different, multimedia data compression it provides a more broad platform, which defines a format for a kinds of frameworks,
embroidered polo shirts, rather than the specific algorithm,
nike air max 2, which hope to build a more free communication and development environment. So the new goal of MPEG-4 is defined as: support a variety of multimedia applications, especially content-based multimedia information retrieval and access, according to different application requirements, site configuration decoder. Coding system is open, feel free to add a new effective algorithm module. Applications include real-time audio-visual communication, multimedia communication, remote monitoring / surveillance, VOD, home shopping / entertainment. MPEG-4 has many advantages. Its compression ratio can be over 100 times and still maintain excellent sound and picture quality; it can use minimal amounts of data to obtain the best image quality, low bit rate
meet the application requirements; it is more AV services for interactive and remote monitoring. In order to meet the needs of a variety of applications, MPEG-4 standard use of very large, with a wide range of adaptability and scalability.
1, shape coding
shape access to information graphics must first be analyzed and divided, each representing different content to the target partition and then use the shape representation. Binary shape information is usually expressed
Alpha plane. Alpha plane is available near the binary arithmetic coding information (CAE); gray Alpha plane motion compensation can be added similar to the texture DCT transform coding to encode the same.
transform for image compression which has discrete Forier transform (DFT), discrete wavelet transform (DWT), singular value decomposition (SVD), K-L transform, Walsh transform, Hadamard transform, Harr transform, Slant transform, discrete cosine transform (DCT). K-L transform them to the best correlation, and DCT is close to the K-L transform results of the most easy to implement the transformation. And MPEG-1 / 2 as, MPEG-4 is also selected DCT. Typically, entropy coding for data compression methods are Huffman (Huffman) coding, vector quantization, arithmetic coding, RLE, LZW coding. For texture encoding, MPEG-4 is selected to run-length encoding, vector quantization and Huffman coding mixed programming code (VLC). Texture coding to go through the DCT transform, quantization, DC / AC prediction, scanning, based on Hufman the VLC coding.
2, motion estimation and compensation
MPEG-4 provides a block-based motion estimation and compensation techniques to effectively use the video content of each VOP time redundancy. General, motion estimation and compensation can be any shape as image sequences for block matching extension. Block-matching process used for the standard macro block; prediction error and used to predict the macro block is encoded with motion vectors; advanced motion compensation support for overlapped block motion compensation mode, 8 × 8 block can be encoded motion vectors. In order to get high coding efficiency of motion estimation, prediction image and the image is more similar to the better prediction, so before the first motion estimation to compensate. MB in the target boundary is then used to fill the first with the level of vertical fill, the rest completely outside the VOP MB with expansion to fill.
3, texture coded
texture refers to the I-VOP image and P / B-VOP by the residual image after motion compensation information. Texture usually compressed transform domain coding and entropy coding. Associate Editor have been published officially: still image compression coding standard (JPEG); store compressed digital audio and video coding standard (MPEG-1); common video compression coding standard (MPEG-2).
then, MPEG Group in February 1999 formally announced the MPEG-4 (ISO/IEC14496) V1.0 version. Version of the same by the end of MPEG-4V2.0 has been completed and officially in early 2000 as an international standard. MPEG-4 standard will be a large number of multimedia applications integrated into a comprehensive framework designed for multimedia communications and applications provide standard algorithms and tools to establish a multimedia transmission can be, storage, retrieval, commonly used in applications such as uniform data format, and according to different application requirements, on-site configuration decoder, open coding system can also join a new and effective algorithm module. To support access to video content, MPEG-4 proposed a
4, flexion and extension of
Currently, MPEG Group has launched a special support for multimedia information retrieval and content-based coding framework of MPEG-7 and multimedia standard MPEG-21. In addition, the ITU-T and the joint development of the new standard MPEG H.264 is the latest video coding algorithms. To reduce the bit rate, get a better image quality as much as possible, H.264 MPEG-4 standard drawing on the strengths, has a higher compression ratio, a better adaptability of the channel, bound in digital video communications and storage fields wide range of applications, and its unlimited potential.
video flexion, including space and time flexion of flexion. Flexion of space can be a different spatial resolution, the time available to flexion of the different time resolution. Each has a multi-flexion, only the high and low in the case of layer 2, the bottom refers to the base layer, while the top refers to the enhancement layer.
5, error avoidance
VLC code a bit error can cause loss of synchronization, and motion compensation would cause error propagation. MPEG-4 to avoid errors in three areas: resynchronization, data recovery and error concealment. Resynchronization refers to the error is detected, the decoder and bit stream re-synchronization between the technology. In general, this method will sync point before the error to the reconstruction of data between the synchronization points discarded. However, these data can be discarded with other techniques to restore and implementation of error concealment. Data recovery tools and the bit stream decoder to re-establish synchronization to restore the discarded data. These tools are not simple code with fault-tolerant recovery, but with a means of avoiding errors that
codewords using reversible VLC VLC coding. Error concealment, in the re-synchronization effectively positioning errors can be easily handled. To further enhance the error concealment ability, it is necessary to increase the error location capabilities, in particular, data partitioning can be used to improve fault localization capabilities.
JVT: a new generation of
JVT video compression standard is ISO / IEC MPEG and ITU-T VCEG Joint Video Team established (Joint Video Team), is committed to a new generation of digital video compression standards.
JVT standard ISO / IEC in the official name: MPEG-4 AVC (part10) standard; in the ITU-T in the name: H.264 (earlier known as H.26L)
H264/AVC
H264 concentration of the advantages of the previous standards, and standards development in the absorption of the previous experience, the use of simple design, making it easier to promote than MPEG4. H.264 creative multi-reference frame, multi-block type, integer transform, intra prediction of new compression technology, the use of more precise sub-pixel motion vector (1 / 4, 1 / 8) and a new generation of loop filter, making the compression performance has greatly improved the system more perfect.
H.264 has the following several advantages:
- efficient compression: H.263 + and MPEG4 SP and compared to 50% bit rate reduction
- delay constraint has good flexibility
- Fault Tolerance
- encoder / decoder complexity scalability
- Decode all the details: no does not match
- high-quality application
- Network Monitoring in a friendly
video coding technology mainly used in the current monitoring MJPEG, MPEG1 / 2, MPEG4 (SP / ASP), H.264/AVC, such as several video coding techniques. Made for the end user to care about him most are: clarity, storage capacity (bandwidth), there is price stability. Using different compression techniques, will significantly affect more than a few major elements.
MJPEG
MJPEG (Motion JPEG) compression technology, mainly developed based on static video compression technology, its main feature is the basic video stream does not take into account the different changes between frames, only one frame is compressed individually.
MJPEG compression technology for high-definition video images, you can dynamically adjust the frame rate, resolution. However, due to not take into account changes between frames, resulting in substantial duplication of redundant information is stored, so a single frame of video take up more space, the popular MJPEG technology can only do the best 3K bytes / frame, usually 8 ~ 20K!
MPEG-1 / 2
MPEG-1 standard is mainly for the SIF standard resolution (NTSC system for the 352X240; PAL system is 352X288) image compression. compression bit rate of the main objectives of 1.5 Mb / s. technology than MJPEG, MPEG1 compression in real time, the amount of data per frame, processing speed has improved significantly. However, there are more negative MPEG1 places: storage capacity is too large, the resolution is not high enough and the network transmission problems.
MPEG-2 in MPEG-1 based on the expansion and upgrading, and backwards compatible with MPEG-1, mainly for storage media, digital television, high-definition and other applications, resolution: Low (352x288 ), medium (720x480), the second highest (1440x1080), high (1920x1080). MPEG-2 MPEG-1 video is relatively improved resolution, high-definition to meet the requirements of users, but not much improved compression performance, making the storage capacity is still too large, but also discomfort and network transmission.
MPEG-4
MPEG-4 video compression algorithms for MPEG-1 / 2 compression at low bit-rate has increased significantly in the CIF (352 * 288) or higher resolution (768 * 576) video compression case, in terms of clarity on or from the storage has more advantages than MPEG1,
jordan shoes on sale, more suitable for network transmission. Also MPEG-4 can be easily adjusted dynamically frame rate, bit rate, to reduce the amount of storage.
MPEG-4 because the system design is too complicated, so difficult to fully implement MPEG-4 and is compatible with, it is difficult in the video conferencing, video telephony and other fields to achieve, which is a little bit from the original in mind. Also for Chinese enterprises have to face high patent fees, at present provides:
- each device needs to decode MPEG-LA 0.25 美元
- encoding / decoding equipment needed payment at the time (4 cents / day = $ 1.2 / month = 14.4 U.S. dollars / year)
H.264/AVC
H.264 standard concentration of the advantages of the past, in many have been breakthroughs in the field,
nike air max 95s, making it get much better overall performance than the standard:
- and H.263 + and MPEG-4 SP savings of up to 50% compared to the bit rate, the storage capacity greatly reduced;
- H.264 at different resolutions and different bit rate can provide higher video quality;
- a more conducive to network transmission.
H.264 using simple design, making it easier to promote than MPEG4 and more easily in the video conferencing, video phone to achieve, making it easier interconnection, can be easily and G.729 low-bit rate voice compression to form a complete system.
MPEG LA's MPEG-4 high absorption of royalties and make it difficult to promote the lessons learned, MPEG LA has established the following fees low H.264: H.264 broadcast basic charge; products embedded in the H H.264 encoder / decoder, the annual output of 10 million units the following do not charge fees, more than 10 each receive $ 0.2 million, more than 500 each receive $ 0.1 million. Low royalty makes it easier for Chinese H.264 monitoring products to the world.
Full HD video encoding
Hybrid Mode
