Computational complexity control for HEVC based on coding tree spatio-temporal correlation

The High Efficiency Video Coding standard shows improved compression efficiency in comparison to previous standards at the cost of higher computational complexity. In this paper, a complexity control method for HEVC encoders based on the dynamic adjustment of the number of constrained coding treeblocks is proposed. The method limits the maximum tree depth used in the coding structures based on spatio-temporal correlation in order to decrease the number of evaluations performed in the Rate-Distortion Optimization process. Experimental results show that the proposed method is capable of maintaining the encoding time per frame under a pre-defined target, reaching computational complexity decreases of up to 50% at the cost of an average BD-PSNR loss of 0.26 dB in the worst case scenario.