An adaptive orthogonal matching pursuit algorithm based on redundancy dictionary

Abstract

The reconstruction performance of an orthogonal matching pursuit algorithm is poor due to less observation values. An observation matrix design method which can adaptively ensure the sample size based on the image information is proposed. To make the algorithm more sparsely representative, an adaptive orthogonal matching pursuit algorithm based on the redundant dictionary is discussed by using a K-SVD dictionary training method to get a sparse dictionary. Experimental results show that the algorithm not only solves the problem that the sample size is small, but also improves the image reconstruction quality.