A novel rate control algorithm of video coding based on visual perceptual characteristic

Abstract

The rate control plays a very important role in video streaming applications. The current rate control scheme in the advance video coding standard H.264/AVC still results in a relatively large quality fluctuation for frame-level with higher motion activities. However, the current video coding standards remove the spatial-temporal redundancy based on the signal statistical correlation. Therefore, to reach higher compression efficiency without perceptually degrading of the reconstructed video signal, the properties of the human visual system need to be better exploited. Research effort has been dedicated to modeling the spatial saliency map based on the properties of HVS. The importance of video image varied with the regions, and human visual attention too varied with the importance of region of video image, so using the attention selectivity properties of the human visual system, the paper proposed video coding rate control method based human visual saliency map in H.264/AVC. For each frame, target bits are allocated through the correction of attention region size between the current frame and previous frame, and for each MB, MB's complexity are modified through the saliency and MAD value of MB. At last, the algorithm was implemented in reference software JM12.2. Simulation results showed that, comparing with traditional algorithm, the proposed reduced the coding bit rate, at the same time, and it also improved the subjective quality of reconstructed video in visual attention region.