Edge detection algorithm for noisy remote sensing image using directional filter

Abstract

Remote sensing image edge detection is the basis of remote sensing image registration and recognition. In view of the noise formed in the process of remote sensing image acquisition, a new edge detection algorithm based on directional filter is proposed to improve the accuracy of remote sensing image registration and recognition. The directional filter is based on wavelet transform denoising, and three main features including scale, direction and shape of wave are introduced. The target edge can be identified more accurately in the noisy image by the directional filter, and the edge detection result image is obtained by binarization segmentation and non-maximum suppression. Through experimental comparison with other denoising detection algorithms, more accurate edge detection results can be obtained, and there are obvious advantages in operation efficiency.