Abstract— Shadow removal from a single image is a challeng-ing problem, whose solution is proposed in this paper using3D intensity surface modeling. Due to the high-order texturalcontent in the original images, a direct modeling of the intensitysurface of shadow image is difficult. In this paper, imagedecomposition technology is used as an edge-preserving filterto remove the textural detail while keeping the local-smoothnesspattern of image intensity surface. Using 3D modeling, a properintensity surface of illumination in shadow region can be obtainedbased on that corresponding to the same texture in the non-shadow one. Thus, the intensity surface of shadow region can becompensated with a respective shadow-removal. Experimentalresults demonstrate the effectiveness of the proposed approachin the aspect of single-image shadow removal. In contrast tothe alternative methods, it is not limited by additional assump-tions or conditions; moreover, it can deal with the non-uniformand curved surface shadows, and is applicable to the shadowregions consisting of different types of textures