Revisiting Local Texture Correlation for Rate-Distortion Optimized Intra Coding

In this paper, we focus on computationally modeling of the local texture correlations, in an effort to better explore the coding modes with higher priorities in the rate-distortion optimized intra coding. In particular, strong correlations and continuities of local texture with neighboring blocks have been revealed in our analysis, and empirical justifications provide us inspirations on the joint optimization of rate-distortion-complexity when angular modes become finer to adapt the local textures. We examine the philosophy with extensive experiments conducted for refining the intra full-RD list. The results show that better coding performance with on average 0.72% and 3.00% BD-Rate savings for the natural scene and screen content sequences can be achieved in AVS3 test model HPM-5.0 under all intra configuration, with negligible encoding and decoding time variations.

1. Introduction

Intra prediction, which takes advantage of the spatial homogeneity to eliminate the redundancies in a single frame, has been a long-standing problem in video coding. The recent development of intra prediction reveals that increasing the number of mode candidates is beneficial for promoting the prediction accuracy. The number of intra prediction modes has been largely augmented from H.264/AVC [1] (9 intra modes) to High Efficiency Video Coding (HEVC) [2] standard (35 intra modes). In Versatile Video Coding standard [3] and the third generation of audio video coding standard (AVS3), in total 67 intra modes are designed [4]. Such fine-grained prediction scheme is capable of effectively capturing the arbitrary texture directions in natural videos, as illustrated in Fig. 1.