一种结合数学形态学和LOG算子的遥感图像边缘检测方法

摘要

提出了一种结合数学形态学和LOG算子的遥感图像边缘检测算法，该算法可有效地解决在混合噪声及复杂细节环境下的遥感图像地物边缘信息的检测。该算法首先利用中值滤波消除原始图像的非高斯噪声，并采用数学形态学进一步消噪及增强地物间的对比度；然后采用LOG（Laplacian of Gassian）算子对处理后遥感图像进行边缘信息检测。实验结果表明，提出的方法能有效地提取遥感图像地物边缘信息。%A combination of mathematical morphology and LOG operator for remote sensing image edge detection is proposed in this paper. It is effective for detecting edge information of remote sensing image under the environment of mixed noise and intricate detail. In the algorithm,median filter is first used for eliminating non-Gaussian noise of original image,and mathematical morphology is further used to eliminate noise and enhance contrast between surface features. Then,LOG(Laplacian of Gaussian) is used to detect edge information of remote sensing processed. Experiment results show that method we proposed can extract surface feature edge information of remote sensing image effectively.