Research of Edge Detection Based on Gabor Wavelet

The basic problem of edge detection is the contradiction between the detection accuracy and the anti-noise performance. In order to solve the contradiction, a new method based on the wavelet transform is proposed. The method reduces image noise by optimizing Gabor wavelet filter and extracts the image edge by using wavelet multi-scale transform. The experimental results show that the method not only effectively suppress noise, but also can get the edge of the image of high-precision parts.