| Video conferencing and video chat are usually faced with a low bit rate application environment.If video stream transmission does not consider the bandwidth,it usually causes video jamming or video content with obvious blockiness.Human eye vision perception is the best measure for the effect of video processing.Therefore,when coding video,you can use the visual characteristics of the human eye to find the visual interest area.Encoding the visual interest area will greatly improve Video encoding effect.This paper presents an efficient region detection algorithm for face region and skin color region in video conference and video chat scene video.It can quickly identify the region of interest during encoding and improve coding efficiency.By improving the encoding algorithm and controlling the size of the video encoding bit rate,the video encoding quality can be improved.This paper proposes a code rate control modeling algorithm based on the region of interest and implements it on the X264 encoding platform.By testing the test sequence and comparing the results,it is proved that the rate control algorithm proposed in this paper does have a very high practical value.This picture's main tasks are:(1)study and analysis the X264 encoder encoding algorithm and rate control algorithm penetrate deeply,to provide the basic guarantee for the next work.(2)From a practical point of view,an efficient region detection algorithm is proposed.Using fast image foreground and background styles,combined with human skin color features to quickly identify areas of interest in the video conferencing and video chat scenes that are of interest to the human eye,namely the face area and the skin color area.(3)A code rate control modeling method based on the region of interest is proposed.By modeling the regions of interest of the human eye,the quantization parameters are estimated and the video quality of the region of interest of the video is improved,thereby improving the overall subjective visual quality.The more accurate rate control and the purpose of improving subjective video quality.(4)The code rate control algorithm based on the region of interest is implemented on the X264 encoding platform,and the test sequence is tested and compared with the results.This article has completed a total of 72 groups of test work,and compares the error rate of the target code rate with the output code rate and the subjective visual effect of the coded video.Among them,in the standard conference scene sequence,the rate error rate of X264 platform is 22.96%,the error rate of this algorithm is 15.36%;in the life education scenario,the error rate of X264 platform is 22.99%,and the error rate of this algorithm is 15.7%;In terms of subjective visual effects,this algorithm is significantly better than the X264 platform. |
PDF Full Text Request