一种基于小波压缩感知的藏族壁画图像处理.

According to the feature of Tibetan murals, a compressed sensing algorithm is presented based on wavelet. The improved compressed sampling algorithm separates the high frequency sub-bands from the low frequency sub-band, and realizes sparse transformation. The compressed sensing OMP algorithm is employed in image reconstruction. Experiments show if the number of blocks and random matrix is fixed, then the sampling rate is higher, the PSNR will be smaller, and the reconstructed image will have better quality. Hence the algorithm is effective for Tibetan murals. [ABSTRACT FROM AUTHOR]