An Orthogonal Matching Pursuit with Thresholding Algorithm for Block-Sparse Signal Recovery

2nd International Conference on Soft Computing and Machine Intelligence, ISCMI 2015

Hong Kong

2016.02.19

In this paper, a block version of the orthogonal matching pursuit with thresholding algorithm is proposed. Compared with the block version of the orthogonal matching pursuit algorithm, the block orthogonal matching pursuit algorithm works in a less greedy fashion in order to improve support estimating efficiency in each iteration. The lower and upper bounds of the threshold are theoretical derived for exact recovering the block-sparse signal in noiseless environment. When the noise is bounded with unknown distribution, the algorithm can also estimate the support exactly under certain ssumptions. Experimental results are provided to illustrate the validity of our main results.