Sparse low-rank correction scalable video coding based on interactive region of interest viewpoint
On the basis of Wyner–Ziv video coding in transform domain, this paper studies the coding method of key frame and proposes a method of coding mode selection based on partition of frequency band: First of all, the key frames are transformed by DCT and they are divided into low-frequency band and high-frequency band according to the characteristics of frequency band. Then the low-frequency band and high-frequency band are encoded and decoded by Wyner–Ziv and intraframe, respectively. In this paper, the decision of coding mode is transferred to the decoder to be completed, which further improves the performance of the system without increasing the complexity of the encoder. Compared with the reference algorithm, the peak signal-to-noise ratio obtained by the proposed coding mode selection algorithm based on partition of frequency band is improved by 1–5 dB.
Journal of supercomputing