An improved method of image edge detection based on wavelet transform

Aiming at the problems of traditional edge detection algorithms such as Sobel, Robert, Canny and Log etc. on noise immunity and detection accuracy, this paper puts forward an algorithm which uses wavelet thresholding method to image denoising based on generalized cross validation (GCV) first. Then go on the edge detection on the image by using two-dimension (2-D) wavelet transform´s based on mult-scale feature and a`trous, selects the edge detection result corresponds to the small-scale. Through the comparison of simulation results between traditional edge detection algorithms and improved edge detection method, this method performs better than traditional edge detection algorithms on detail reserving and positioning accuracy.