Abstract: In order to catch the edge information of high resolution remote sensing image more effectively, a new method to get image edge was proposed. Firstly, the main information was collected by Principal Component Analysis (PCA) transform. Secondly, the information was divided with symletsA wavelet while the image in each scale was processed with morphological operators. Finally, the edge of image was enhanced by implementing correlation filtering to the image in the same scale with filtering algorithm of wavelet phase, and the edge information was caught by partitioning the image with OTSU algorithm. The results show that, compared with the existing algorithms, the edge of image is located more accurately and the edge detection effect is more evident with this method.

Key words: sensing image, symletsA wavelet transform, morphology, wavelet phase filtering, edge detection

摘要： 为了更好地获取高分辨率遥感影像的边缘信息，提出一种新的影像边缘检测方法。该方法首先利用主成分分析(PCA)变换获取影像的主要信息；然后采用symletsA小波对其进行分解，并用形态学算子对各尺度影像进行处理；最后利用小波相位滤波算法在同一尺度上进行相关性滤波以增强图像边缘，并通过OTSU算法进行分割获取其边缘信息。结果表明：与现有算法相比，该方法对边缘的定位更加精确，边缘检测效果更加明显。

关键词: 遥感影像, symletsA小波变换, 形态学, 小波相位滤波, 边缘检测