HEVC intra partitioning and mode decision using histograms of oriented gradients

A constant need for higher encoding efficiency is fulfilled by the newest video coding standard H.265/HEVC. It was designed to support higher resolution video like FullHD and 4K, but the cost brought by the improved coding efficiency is a major increase in computational complexity. This paper proposes a new algorithm for intra mode decision in HEVC, targeting real-time video processing. Our method achieves a considerable complexity reduction in intra prediction. It reaches the optimal coding-unit size and angular mode decision with a pre-processing step designed to evaluate the texture. Using an adjusted histogram of oriented gradients, the algorithm selects the best choice in CU size and computes an estimate for the dominant edges. Testing performed on two test video sequences showed significantly lower encoding times for the same quality, compared to the original reference software.

[1]  Gary J. Sullivan,et al.  High Efficiency Video Coding (HEVC), Algorithms and Architectures , 2014, Integrated Circuits and Systems.

[2]  Luciano Volcan Agostini,et al.  Fast HEVC intra prediction mode decision based on EDGE direction information , 2012, 2012 Proceedings of the 20th European Signal Processing Conference (EUSIPCO).

[3]  Yong-Jo Ahn,et al.  A context-adaptive fast intra coding algorithm of high-efficiency video coding (HEVC) , 2016, Journal of Real-Time Image Processing.

[4]  Bill Triggs,et al.  Histograms of oriented gradients for human detection , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[5]  Gaobo Yang,et al.  Fast CU size decision and mode decision algorithm for intra prediction in HEVC , 2015, Multimedia Tools and Applications.

[6]  Farzad Zargari,et al.  Two stage fast mode decision algorithm for intra prediction in HEVC , 2015, Multimedia Tools and Applications.

[7]  Luciano Volcan Agostini,et al.  Fast mode selection algorithm based on texture analysis for 3D-HEVC intra prediction , 2015, 2015 IEEE International Conference on Multimedia and Expo (ICME).