Abstract:

Compressed Sensing(CS) can reduce the cost to measure a signal by using its characteristic as sparse in wavelet domain.So CS can break the limit of Nyquist sampling rate.For overcoming the shortage of orthogonal wavelet bases which has poor directional selectivity and week ability to capture the image edge,Image Compressed sensing based on Bandelet transform domain is proposed.The Bandelet algorithm can track the regular direction of image adaptively to improve image sparse ability.,thereby improving the quality of reconstruction image.Experiment results show that the proposed algorithm protects image details and improves the peak signal-to-noise ratio and visual quality.

Key words: compressed sensing;image sparse;bandelet;wavelet transform