An image edge is usually complicated in complex background. An ear-edge-extraction method based on wavelet modulus maxima is investigated in this paper in order to solve the contradiction between the requirement of regional precision and the property of anti-noise in the process of edge-detection. Firstly,the images of wavelet modulus maxima at three different scales are obtained after image wavelet decomposition. Secondly,these images are separately transformed to binary edge images which are superposed then. Thirdly,noisy points outside the skin color region are excluded by using the binary image of skin color and then the edge image is obtained. After morphological processing on the original image,the binary image of skin color can be obtained with prior knowledge of side face. In this method,the technology of image superposition based on wavelet modulus maxima of multi-scale is exploited to solve the problem that the ear features of internal and external edges are hard to be extracted. Experimental results show that this method is effective in complex background.