Subjective Rate Distortion Optimization for Screen Content Coding Based on SVR
Haiying Peng, Shun Hu, Tong Tang · 2021 13th International Conference on Wireless Communications and Signal Processing (WCSP) · 2021
Screen content video is widely used in wireless display, remote desktop, video conference, etc. The massive data generated by these applications brings great pressure to transmission and storage, so it is an urgent need for efficient screen content coding (SCC) method for data compression. In this paper, a subjective rate distortion optimization method based on SVR is proposed for SCC, which significantly improves the compression efficiency of screen content. Firstly, we extract the structure texture feature set of the screen content image based on the visual characteristics. Then, we filter the appropriate texture features as the feature vector of SVR, and use SVR to train the optimal regression model. Finally, the regression model is used to predict the Lagrange multiplier of the coding block and improve the rate distortion optimization process. Experimental results show that compared to state-of-the-art method, the proposed method can save up to 9.46% bitrate cost under similar subjective quality.