|
Water body detection in high-resolution aerial images is quite different from that in low-resolution aerial images because of the color, shape and size varieties of water bodies. In this paper, an automatic water body detection method based on stereo processing and analysis of stereo aerial images is proposed. The special color and disparity distribution pattern of pixels in the water bodies are studied through wavelet transform analysis. Furthermore, a score based on the above distribution pattern is proposed to classify the water bodies from other terrestrial features. Experimental results show the effectiveness of the proposed method in detecting water bodies of various color, shape and size efficiently.
|
