In this paper, we propose a denoising algorithm for remote sensing images based on sparse representation and inter-scale correlation model. First discrete wavelet transform is used to decompose the remote sensing images. The low-frequency coefficients are analyzed by training the over-complete dictionary with K-SVD algorithm, and the high-frequency coefficients by correlation analysis. The correlation quantity of high-frequency coefficients at each scale is calculated and compared with the high-frequency coefficients at the corresponding positi...
