The Metal Surface Defects Detection Algorithm Of Edge Detection Research And Application

Edge plays one of important roles in the image, and the edge detection is the base of the computer vision and pattern recognition. Because of the noise in the image signal and all kinds of fuzzy edge combined in the image, the edge detection is one of the difficult problem in the image processing. Both the noise and the edge belong to high frequency signal, it's hard to pick up with frequency band. Wavelet can replace any one method using by Fourier. Wavelet transform has the character of math microscopy, it's the challenge to the Fourier. In the image processing, the scale of the wavelet can adaptively adjust and detect according to the difference between the character in the local image. Research on image detection based on wavelet makes up the disadvantage in the traditional edge detection and improves the edge localization precision avoiding noise effectively.The research of Image Edge detection is developing deeply, and become one of hot spots of study. This paper will discuss the problem of Image edge detection. An improved edge detection method based on wavelet transform and image fusion is brought forwards. The image is decomposed into high frequency and low frequency parts using wavelet transform, and edge detection is made with them respectively. Image fusion between high frequency edge and low frequency edge is done using local area square-difference principle in wavelet region. The experiment shows that the method can fuse character of high frequency and low frequency edge image, and have a better function of edge detection.