Abstract: An effective water body extraction method for high-resolution remote sensing images is proposed. Firstly, a local binary patterns and [K]-nearest neighbor based approach is designed to classify the water body and the ground coarsely, and the morphological filter is adopted to suppress the noises in the divided regions. Secondly, a local binary patterns and support vector machine based approach is proposed to finely classify the boundary regions between the water body and the ground and at the same time, the morphological filter is used to suppress the noises in the boundary regions. Finally, the morphological erosion operator is applied to the refined results so as to smooth the edges of the water body. The experimental results show that the proposed algorithm can efficiently and effectively detect the water body targets in the remote sensing images.

Key words: high-resolution remote sensing image, water body extraction, local binary patterns;[K]-nearest neighbor, support vector machine

摘要： 针对高分辨率遥感图像，提出一种有效的水体自动提取算法。设计基于局部二值模式和[K]最近邻的算法，对遥感图像进行水陆粗分离，并采用形态学滤波抑制各分类区域内的噪声；设计基于局部二值模式和支持向量机的方法，对水陆边界区域进一步细分离，并对细分离结果采用形态学滤波去除水陆边界附近噪声点；针对细化结果，采用形态学腐蚀运算对水体边缘进行平滑，得到最终的水体提取结果。实验结果表明，提出的算法能快速、有效地检测遥感图像中的水体目标。

关键词: 高分辨率遥感图像, 水体提取, 局部二值模式, [K]近邻, 支持向量机