Compressed video capture and reconstruction system based on structured sparse dictionary learning

Abstract

The invention provides a compressed video capture and reconstruction system based on structured sparse dictionary learning. The system comprises a structured sparse dictionary learning module, a video signal sensing module and a reconstruction processing module. The structured sparse dictionary learning module firstly acquires a training set through a sub-space clustering method, then, a dictionary is acquired through a linear sub-space learning method and minimized-block-relevant block sparse dictionary learning method, the sensing module projects video signals in an image block mode, and acquired data are finally decoded and reconstructed in the reconstruction processing module. Compressed sampling is provided, the distributed progressive structure of the video sampling process is combined, the reconstruction accuracy and efficiency are improved for the special structure of a structured sparse dictionary matrix, the sampling efficiency of the video signals is greatly improved, reconstruction gains are acquired compared with other methods under different sampling compression ratios, and meanwhile the good expandability is achieved.

Description

technical field [0001] The invention relates to a video signal acquisition scheme, in particular to a compressed video acquisition and reconstruction system based on structured sparse dictionary learning. Background technique [0002] The acquisition and encoding (compression) of video signals is crucial for applications such as storage and transmission of video. The traditional signal processing system adopts the mode of sampling first and then compressing: in order to completely preserve all information of the signal, the video should be sampled at a sampling frequency not less than twice the signal bandwidth; the collected original signal is removed after a series of encoding techniques For the purpose of redundancy, the bottleneck of related technologies is that a large amount of sensors and computing resources are spent to obtain a small amount of signal compression data after processing, and the demand for resources at the sampling end is too high. In order to further...

AI Technical Summary

Problems solved by technology

[0003] After searching the literature of the prior art, it was found that "A Theory for Sampling Signals From a Union of Subspaces" published by Yue M.Lu and Minh N.Do in the journal "IEEE Transactions on Signal Processing" (TSP) in 2008 In this paper, a signal sampling theory based on subspace set is proposed. This theory gives the uniqueness and stability conditions to be satisfied for the sampling of the signal in the subspace set, but the subspace set assumed by the theory is composed of fixed Based on Zhangcheng, it cannot provide more effective sparsity and adaptability

In the article "Union of Data-driven Subspaces via Subspace Clustering for Compressive Video Sampling" published at the "IEEE Data Compression Conference" (IEEE DCC) conference in 2014, Y.Li and H.Xiong proposed a data-driven subspace set based on The model applies compressed sensing to video sampling. This method directly compresses and samples the video signal at the sampling encoding end, and uses the UoDS base as a sparse base to reconstruct the signal at the decoding end. This method can flexibly and effectively analyze the signal. Sparse representation is used to ensure the subjective quality of the reconstructed video, but the UoDS base used in this method does not consider the overlap between the subspaces, and the correlation between blocks is so high that it is impossible to obtain a compact block sparsity , thereby reducing the effect of

Images