Compressed sensing based on the improved wavelet transform for image processing

Abstract

Compressed sensing theory is a new sampling theory that can sample signal in a below sampling rate than the traditional Nyquist sampling theory. Compressed sensing theory that has given a revolutionary solution is a novel sampling and processing theory under the condition that the signal is sparse or compressible. This paper investigates how to improve the theory of CS and its application in imaging system. According to the properties of wavelet transform sub-bands, an improved compressed sensing algorithm based on the single layer wavelet transform was proposed. Based on the feature that the most information was preserved on the low-pass layer after the wavelet transform, the improved compressed sensing algorithm only measured the low-pass wavelet coefficients of the image but preserving the high-pass wavelet coefficients. The signal can be restricted exactly by using the appropriate reconstruction algorithms. The reconstruction algorithm is the key point that most researchers focus on and significant progress has been made. For the reconstruction, in order to improve the orthogonal matching pursuit (OMP) algorithm, increased the iteration layers make sure low-pass wavelet coefficients could be recovered by measurements exactly. Then the image could be reconstructed by using the inverse wavelet transform. Compared the original compressed sensing algorithm, simulation results demonstrated that the proposed algorithm decreased the processed data, signal processed time decreased obviously and the recovered image quality improved to some extent. The PSNR of the proposed algorithm was improved about 2 to 3 dB. Experimental results show that the proposed algorithm exhibits its superiority over other known CS reconstruction algorithms in the literature at the same measurement rates, while with a faster convergence speed.