In order to extract water information from high-spatial remote sensing images, wavelet technique was used to expend the information of the object images, and clean main noise of the images; then another method was present to conserve linear water, which is called multi-window linearity reserve technique, then character noise cleaning is used to clean the image noise.

This paper proposed a new mode to substitute the representation to image contents by image union feature for single image feature and then introduced the indexing method of spatial data structure after defining the similarity of images in the Euclidean space.

Experimental result shows that the Combined Denoising Method(CDM) denoises effectively in both homogeneous areas and areas with edges,suppresses artifacts evidently and has high peak signal-to-noise ratio.

Based on the stacking of data matrix,real-processing of signal subspace,complex shift invariance equation and joint eigenvalue estimation,an algorithm is presented to solve this problem and handle the estimation jointly.

The problem of joint eigenvalue estimation for the non-defective commuting set of matrices A is addressed.

To improve the segmentation performance,a novel approach for shadow detection integrating multiple features was proposed,which considers information of color,shading,texture,neighborhoods and temporal consistency to detect shadows in a scene.