Shadows exist in images where three-dimensional objects are irradiated by strong directional lights. Shadows will significantly influence the effects of image processing, such as object detection and image segmentation. Shadow removal has been an important subject as one kind of preliminary processing methods and varies algorithms have emerged. In this paper, we present an algorithm of shadow removal for single texture regions using histogram matching to detect shadows and color model recovery to remove shadows. Results of the whole algorithm show good performances in related experiments. And effects of the intermediate process are also presented in this paper.