A generally analyzed H.264 format before the first five or six years, not very detailed, can look substantially what formats. H.264 Code Flow Structure Analysis
At that time, the Baidu library was uploaded. I used to remember how to be, I can't download it now, but I have to pay for money. Really ~~~
1. Introduction to H.264
MPEG (Video Coding Experts Group) has been jointly developed a better video compression coding standard than the MPEG and H.263 performance than early development, which is named AVC (Advanced Video Coding). Also known as ITU-T H.264 recommends and the standard of Part 10 of MPEG-4, referred to as H.264 / AVC or H.264. This international standard has been officially promulted by ITU-T in March 2003 and officially promulgated internationally. In order to adapt to the demand of HD video compression, in 2004, the Frext section was added; in order to adapt to different code rates and quality, the scalable encoded SVC was added in 2006.
2. H.264 encoding format
The code stream structure defined by H.263 is a hierarchical structure, a total of four layers. Since the top and bottom, the image layer (a gob layer), a macroblock layer, and a block layer (Block Layer). Compared with H.263, the code stream structure of H.264 and H.263 have a great difference, which is no longer a strict hierarchical structure.
H.264 supports encoding and decoding of continuous or interlaced video of 4: 2: 0. H.264 compression is doubled compared to H.263, MPEG-4, and the video compression ratio is doubled.
The function of H.264 is divided into two layers: video coding layer (VCL, Video Coding Layer) and Network Extract (NAL, Network Abstract Layer). The VCL data is the output of the encoding process, which represents the compressed encoded video data sequence. These encoded VCL data are first mapping or encapsulated before VCL data transmission or storage. Each NAL unit includes an original byte sequence load (RBSP, RAW BYTE SEQUENCE PAYLOAD), a set of NAL headers corresponding to video encoding. The basic structure of RBSP is: filling the end bit behind the original coded data. A bit "1" several bits "0" so that the byte is aligned.
Figure 1 NAL unit sequence
3. H.264 transmission
The encoding video sequence of H.264 includes a series of NAL units, each NAL unit contains an RBSP, shown in Table 1. The encoder (including the data split sheet IDR) and the sequence RBSP ending match is defined as the VCL NAL unit, and the rest is NAL unit. A typical RBSP unit sequence is shown in Figure 2. Each unit is transmitted in a stand-alone NAL unit. The information head (one byte) of the unit defines the type of the RBSP unit, and the rest of the NAL unit is RBSP data.
Figure 2 RBSP sequence example
Table 1 RBSP description
4. H.264 code stream structure
Figure 3 H.264 Code Straightening Structure
Start Code: If the SLICE corresponding to the NALU is the beginning of a frame, use 4 bytes, ie 0x00000001; otherwise use 3 bytes, 0x000001. Nal header: forbidden_bit, nal_reference_bit (priority), NAL_Unit_type (Type). Shell operation: In order to make the NALU body do not include the start code, one byte 0x03 is inserted in the encoding of the two bytes (continuous) 0, inserted a byte 0x03 to distinguish between the start code. When decoding, the corresponding 0x03 is removed. NALU type:
Table 2 Nal_Unit_type semantics
5. H.264 decoding
NAL_REFERRENCE_IDC (NRI) for NAL_REFERRENCE_IDC (NRI) is used to mark an importance of a NAL unit during reconstruction. The value is 0 indicates that this NAL unit does not use prediction, so it can be discarded by the decoder without error diffusion; value is higher than 0 Indicates that the NAL unit is used for no drift reconstruction, the higher the value, the greater the loss of the NAL unit loss. The hidden bit of the NAL header information is defaulted in the H.264 encoder. When the network recognizes that there is a bit error in the unit, it can be set to 1. The hidden bit bit is mainly used to adapt to different types of network environments (such as a wired-wireless environment).
Figure 4 NAL unit decoding
The process of NAL unit decoding is: First extract the RBSP syntax structure from the NAL unit, and then processes the RBSP syntax structure in accordance with the flow as shown in FIG. The NAL unit is input, and the output result is a sample point of the current image that is decoded. The sequence parameter set and the image parameter set are included in the NAL unit. The image parameter set and sequence parameter set is used in other NAL unit transmission, and in the tape of these data NAL units, the image parameter set number used by the syntax element PIC_PARAMETER_SET_ID is set; and each image parameter is concentrated in each image parameter. Set the sequence parameter set numbers they use by syntax elements SEQ_PARAMTER_SET_ID.
6. Syntax elements of each hierarchy reference G50 standard.
Our other product:
