Structural Quality Assured Global Optimization for CTU-Level Rate Control of Screen Content Coding
Tong Tang, Yuan Tan, Shihang Ding, Zhidu Li · 2023
Screen content video (SCV) has been applied to various scenes, e.g. online meeting, remote desktop, cloud game, AR/VR and so on, generating massive screen content data and bringing huge pressure on transmission bandwidth. Rate control is an effective technique for improving bandwidth utilization, but existing rate control methods of SCV has not exploited characteristics of human visual system (HVS) fully, leaving space to improve coding efficiency. Therefore, this paper studies the characteristics of HVS and proposed a subjective rate control method for screen content coding (SCC). Concretely, first, a novel coding tree unit (CTU) level subjective rate-distortion (R-D) model is built by utilizing structural feature. Then, on this basis, a global optimization scheme of CTU-level bit allocation is derived. Finally, experimental results show that compared to the default rate control method in HM-16.20+SCM-8.8, the proposed method could averagely save 2.64% bit cost under similar subjective quality.