Selective motion estimation strategy based on content classification for HEVC screen content coding
Mengmeng Zhang, Shuai Wang, Bin Li · 2017
Motion estimation is one of the most time-consuming parts in video coding. This paper proposes a selective motion estimation strategy to reduce the complexity of the encoder for HEVC screen content coding. The proposed scheme is based on content classification. For each screen content inter frame, it is classified into screen content areas and camera-captured areas. Instead of using one single motion estimation search method for the whole frame, different methods are adaptively switched for different areas. Experimental results show that, compared with SCM-7.0 anchor, the proposed strategy achieves 15% and 13% encoding time saving on average merely with average 0.20% and 0.12% Y/G BD-rate loss for low delay and random access configurations, respectively.