Abstract: The sharp variation of gray levels of an image, measured at different scales, can be detected from the local maxima of its wavelet transform. In this paper we describe an algorithm that can extract the edges of an image. The model is based on the two-dimensional dyadic wavelet transform. It provides information of the image boundaries by determining the local maxima of the two-dimension wavelet transform of an image at dyadic scale. Experimental results shows that this method is quite effective.

Key words: image edges, two-dimension wavelet transform, local maxima