In comparison with the standard RGB or gray-scale images the usual multispectral images (MSI) is intended to convey high definition and an authentic representation for real world scenes to significantly enhance the performance measures of several other tasks involving with computer vision, segmentation of image, object extraction and object tagging operations. However, in practices a MSI image is always prone to corruption by various sources of noises while procuring the images. In this paper we propose a methodology which effectively denoised the MSI images through a novel Decomposable Pixel Component Analysis Algorithm. The proposed algorithm uses two essential properties of a MSI image to its advantage, i.e., the dissimilarity of nonlocal space and the global illumination of the correlated regions across wide spectrum; to consider the correlation between non-variant local sub-spaces within the given noise influenced MSI image. The spatial regularity of neighboring pixels within the image is maintained while the block's homogeneity is checked to extract the decomposable pixels which latter is replaced by neighboring full band patches. Keywords: Multi spectral image denoising, spatial regularity, satellite imagery, noise removal
