HEVC (High Efficiency Video Coding) inter-frame prediction fast algorithm based on RC (Rate Control) prediction

Abstract

The invention discloses an HEVC (High Efficiency Video Coding) inter-frame prediction fast algorithm based on RC (Rate Control) prediction, and belongs to the field of source coding communications. The coding depth is determined by comparing residual values, and motion regions are divided through motion intensity indexes; different candidate prediction modes are corresponding to different motion regions, unnecessary inter-frame prediction modes in different motion regions are excluded, and an unnecessary mode determination process is reduced, so that the purpose of reducing the motion complexity is achieved, and the coding efficiency is further improved.

Description

technical field [0001] The invention belongs to the field of information source coding communication, in particular to an HEVC inter-frame prediction fast algorithm based on RC prediction. Background technique [0002] In the rate control (RC) stage, the video encoder encodes and compresses the original data to obtain as little reconstruction distortion as possible or to obtain the lowest possible bit rate. To this end, many new technologies are used, such as more complex inter / intra prediction algorithms, variable block size motion compensation, multi-mode block division, variable size block transformation, motion vectors supporting 1 / 4 pixel precision, and rate-distortion optimization techniques. Wait. These improvements in compression performance are at the expense of increasing a large amount of computation, which brings great inconvenience to real-time video coding and communication. [0003] The International Standards Organization ITU-T's Moving Picture Experts Grou...

AI Technical Summary

Problems solved by technology

Some use image texture information and stable area detection information to determine whether the current CU needs to be further divided by setting a threshold value; some use the rate distortion (RD) cost to determine the threshold value to terminate the current CU in advance. However, the selection of the inter-frame prediction mode has not been effectively processed; some use the distribution characteristics of the gray level difference to segment the current block, but the distortion of the gray level binarization filter is large, and the effect is not ideal; some use The optical flow method calculates the eigenvalues ​​of the Pyramid Motion Divergence (PMD) to determine the segmentation of the CU, which effectively reduces the complexity, but does not fully consider the correlation between motion vectors, resulting in poor performance of the encoding RD; some use the JND model and The ADD decision quickly selects the coded depth and inter-frame prediction mode, but for sequences with complex textures, the algorithm has limited effect on reducing complexity; some predict depth values ​​by weighting adjacent CUs in the spatio-temporal domain, reducing the maximum coding unit ( The number of depth traversals of LCU), but its fixed weight algorithm does not consider the differences between sequences, resulting in errors in its predicted depth range

Images