Improvement of Shadow Detection Algorithm Based on c₃ Component for High-resolution Remote Sensing Images

This paper proposes an improved shadow detection algorithm based on c₃ component for remote sensing images, aiming at the sensitivity of c₃ to noise and the inaccuracy of the available algorithms of c₃ component. Use of logarithmic transformation and thresholding method in image enhancement improves the contrast ratio of shadowed to non-shadowed area, thus increasing the recognizability of shadowed pixels. Neighborhood was extended and weights of influential pixels were increased for image smoothing, which decreases noise effect on shadow extraction. By adding operators in two directions and selecting edge threshold based on Gaussian distribution and sampling method, accuracy of shadow edge detection is improved. Experimental results show that the proposed method increases the accuracy of shadow detection and has a strong adaptability, thus can be well applied into the process of high-resolution remote sensing images.