Abstract:This paper mainly study the edge extraction algorithm of thenar palmprint image. Some classical edge detection operators and Hough transform are introduced. The wavelet modulus maxima multi-scale edge detection method is highlighted. A multi-scale Gaussian edge detector is constructed. According to the characteristics of cross-scale transfer of wavelet transform modules for the noise and the image edge, the method of wavelet modulus maxima multi-scale edge detection is adopted to extract the edge of the thenar palmprint image. The experiment results show that this algorithm can reduce the effects of noise and detect rich edge details; meanwhile, the position of fused edges is precise. Whereas, it has the deficiency of poor continuity at the edges.