Code performance improvement scheme for X264 based on SSIM

X264 takes objective metric as distortion criteria to perform Rate Distortion Optimization (RDO)-based mode decision, which cannot acquire optimal subjective quality. This paper applies Structural Similarity (SSIM) based subjective distortion to RDO-based inter mode decision in X264 video coding, and puts forward a method of adaptive analysis of Lagrange multipliers to measure the relation of rate and SSIM in macro-block level. We also take motion vector of macro-block into account and apply methods of information theory, combined with the uncertainty of visual speed perception, to make further improvements of lagrange multiplier. Our experiments show that this correct solution gets a better subjective rate-distortion performance and subjective image quality.