Abstract
Versatile Video Coding (VVC) has advantages over High Efficiency Video Coding (HEVC); it can save nearly half of the bit rate and significantly improve the compression efficiency, but VVC’s coding complexity is extremely high. Therefore, VVC encoders are difficult to implement in video devices with different computing capabilities and power constraints. In this paper, we apply texture information and propose a VVC intra complexity control algorithm. The algorithm assigns a different encoding time to each CU based on the corresponding texture entropy. Besides, the complexity reduction strategy at the CU level is designed to balance the complexity control while taking rate-distortion performance into consideration. Experiments in our paper show that the coding complexity can be accurately controlled from 90% to 70% with a slight loss of RD performance.
Funding Information
  • Natural Science Foundation of Chongqing (cstc2021jcyj-msxmX0411)
  • Natural Science Foundation of Zhejiang Province (LY20F010005)
  • Science and Technology Research Program of Chongqing Municipal Education Commission (KJZD-K202001105)
  • Scientific Research Foundation of Chongqing University of Technology (2020zdz029, 2020zdz030)

This publication has 31 references indexed in Scilit: