A Novel Post-Processing Algorithm for Parallel-Decoded U-HDTV Video Sequences

Abstract

Ultra-High Definition Television (U-HDTV) promises to offer a strong sense of reality to viewers by utilizing extremely high spatial and temporal resolution. In order to compress U-HDTV video sequences in a practical way, the video sequence is divided into sub-sequences and each sub-sequence is processed. However, the merged sequence of independently decoded sub-sequences inherently exhibits a distortion along the boundary of the sub-sequences, namely sub-sequence boundary distortion (SBD). In this paper, we first introduce a method to quantitatively measure the SBD. Then, we propose a post-processing algorithm that reduces the SBD. Experimental results indicate that the SBD is efficiently alleviated by the proposed post-processing algorithm.(1)