The shadow areas have the characters of lower brightness, higher saturation and greater hue value. Over-segmentation of the high resolution remote sensing image was operated, and the red, green and blue color image was transformed to the hue, intensity and saturation space. A shadow index was calculated using the hue, intensity and saturation, which could separate shadow from other classes. An object-oriented processing was performed on the shadow index results, which could keep the spatial geometry information of image. OSTU algorithm is used to choose a suitable threshold value. With the threshold value, the shadow detection results could be acquired by the histogram thresholding segmentation algorithm. Experimental results demonstrate that the proposed approach can be used to separate shadow from other classes effectively.