A Video Coding System with Spatial-Temporal Down-/Up-Sampling and Super-Resolution Reconstruction

摘要

A video coding system based on spatial-temporal down-sampling/up-sampling and super-resolution reconstruction method is proposed. The method is an exploration and providing ideas for our future work. The system pre-reduces the spatial resolutions and frame rates of source videos before being encoded by a standardized codec such as HEVC/H.265. On the reception side, the decoded video is up-sampled to four times resolution, and the frame rate is also increased to original frame rate. We tested the performance in combination with a 2K 60P HEVC/H.265 encoder. The experiment results are shown in this paper.