A new method for noise estimation in single-band remote sensing images
Peng Fu, Quansen Sun, Zexuan Ji, Qiang Chen · 2012
In this paper, we proposed a new method for noise estimation in single-band remote sensing images. The new algorithm constructs the intensity-homogenous blocks by utilizing the high-pass operators with eight directions, and estimates the noise by using spatial de-correlation via multiple linear regression. The final noise estimation result can be automatically calculated. We compared our algorithm to other block-based estimation approaches in both artificial and real remote sensing images. Experimental results demonstrate that the proposed algorithm is more robust and stable, and can produce more accurate and reliable estimation results for the images with complex edges and textures, especially for real remote sensing images when noise level is not too high.