Joint Temporal-spatial Domain for Adaptive Block Compressed Sensing Video Reconstruction Algorithm

摘要

An adaptive block-based compressive sensing (BCS) video reconstruction algorithm based on temporal-spatial domain characteristics is proposed. Firstly, assigned weight function is introduced, and the adaptive sampling scheme of joint wavelet coefficients and variance are developed. On the basic, the global measurement matrix is constructed by assigned weight matrix to realize global reconstruction. Secondly, combined with the block-based multi-hypothesis (MH) model and the minimum total variation (TV) model, the predictive-residual reconstruction model of joint temporal-spatial domain characteristics is constructed, and the prediction frame of the current frame is obtained by iteration. Thirdly, the residuals are calculated by global reconstruction and combined with the predictive frames of the current frame to reconstruct a new frame, To validate the effectiveness of proposed video reconstruction algorithm, the results are compared with other BCS algorithms which proposed from the aspects of video reconstruction in recent years. The experiment results show that the proposed algorithm can effectively improve the quality of video reconstruction and further reduce the computational complexity compared with other algorithms.