Rate-Distortion Cost Estimation Model Based on Cauchy Distributions for HEVC Encoder

In High Efficiency Video Coding (HEVC), Rate-distortion optimization (RDO) process aims to select the optimal coding parameters according to a certain strategy in order to achieve the optimal coding performance, and therefore leads to an increase of computation complexity. For this, single probability model approaches have been taken in previous studies due to their simplicity but are not popular in HEVC because of the quadtree-based coding structure. Mixture probability models have better accuracy but they also result in computation complexity. So this paper proposed a new rate-distortion cost estimation model based on Cauchy distribution with a constantly revised parameter which has low complexity and proved to have high accuracy.