An Improved Compressed Sensing Method for High Resolution Image Denoising

As image resolution becomes higher and higher, high resolution image processing has become an urgent technological problem. To further improve the quality of reconstructed high resolution image, an improved compressed sensing (CS) iterative thresholding algorithm is proposed. First, an optimized wavelet threshold is proposed to select the image coefficients in each iteration; Then, an improved total variation (TV) method is proposed to adjust the reconstructed image for further improving the reconstructed image quality. The experimental results show that the algorithm proposed can recover a high quality image while preserving more detailed information. When compression sampling ratio is lower, the proposed algorithm still can obtain correspondingly higher peak signal to noise ratio (PSNR).