Novel HEVC (High Efficiency Video Coding) rate control GOP (Group Of Pictures)-level bit allocation method

Abstract

The invention discloses a novel HEVC (High Efficiency Video Coding) rate control GOP (Group Of Pictures)-level bit allocation method, and belongs to the technical field of video coding. Defects of an existing HEVC rate control GOP-level bit allocation method are discovered through in-depth study for a coding structure of HEVC, and a novel HEVC rate control GOP-level bit allocation method is provided. According to the invention, bits of each GOP which does not contain an I-frame in every I-frame intra period are enabled to be uniformly allocated, and the rate distortion performance is improved. In addition, the state of a buffer area is enabled to return back around 0 at the end of coding of each I-frame intra period, so that the rate control precision is improved, and the size of a buffer area required in coding and the delay in transmission are reduced.

Description

technical field [0001] The invention belongs to the field of video coding, and in particular relates to a new HEVC (High Efficiency Video Coding) code rate control GOP-level bit allocation method. Background technique [0002] With the rapid development of communication technology and multimedia technology, video applications have penetrated into every aspect of people's lives. Bit rate control is a key technology in video coding. Its goal is to make the bit rate after video coding meet the required bit rate limit and make the distortion as small as possible during coding. Rate control usually consists of two steps: target bit allocation and quantization parameter determination. For HEVC code rate control, the target bit allocation generally adopts a hierarchical bit allocation strategy, and the target bits are allocated to the coding units at the GOP (Group Of Pictures) level, frame level, and CTU (Coding Tree Unit) level in turn; the second step of rate control It is the...

AI Technical Summary

Problems solved by technology

[0004] (1) Choi and Yoo transplanted the rate control method of H.264 to HEVC, specific reference "H.Choiand J.Yoo, "Pixel-wise unified rate-quantization model for multi-level RateControl," IEEE Journal of Selected Topics in Signal Processing, vol.7, no.6, Dec2013.", but this solution does not consider the difference between the GOP structure of HEVC and the GOP structure of H.264, so that the bit allocation of the GOP behind the GOP containing I frames Too little will seriously affect the normal allocation of bits and coding performance

[0005] (2) Wang and Ma et al. distribute bits evenly at the GOP level. For details, refer to "S.Wang and S.Ma, "Rate-GOP based rate control for High Efficiency Video Coding," IEEE Journal of Selected Topics in Signal Processing, vol.7, no.6, Dec 2013.", but this scheme does not consider the difference between the GOP structures under the open-GOP structure under the HEVC random access (Random Access) configuration, so that the random access configuration does not contain I-frame GOP allocates too many bits

[0006] (3) Li Bin and others proposed a novel λ-domain code rate control method, the specific reference "B.Li and H.Li, "λdomain rate control algorithm for High Efficiency Video Coding," IEEETransaction on Image Processing, vol.23 , no.9, Sep 2014.", but in order to smooth GOP bit fluctuations, this method uses a sliding window mechanism to smooth the GOP-level bit allocation method

However, under the random access configuration, such as figure 1 As shown, the sliding window mechanism cannot make the encoding buffer fullness return to 0 at the end of an I frame insertion period (Intra period) encoding, and the encoding buffer status stays at a high level for a long time, resulting in a relatively large buffer required for encoding and decoding. large; and make in an I frame insertion cycle, the bits that the GOP that affects the bigger before are distributed are few, and the bits that affect the smaller GOP that affects the back are many (such as figure 2 shown), affecting encoding performance

In the Low Delay (Low Delay) configuration, there are similar problems, image 3 It is the encoding buffer state curve diagram of the λ-domain code rate control algorithm under the low-delay configuration. Figure 4 This is the GOP bit fluctuation curve of the λ-domain code rate control algorithm under the low-delay configuration. The GOP bits behind the I frame show an increasing trend, which affects the coding performance.

Images