Abstract: An algorithm used to detect aerial image's edge contaminated by addictive white noise is discussed. Using statistical characteristic of contaminated aerial image's wavelet coefficients and multi resolution decomposition of wavelet transform, shrinkage factors are constructed and used for treating the edge or noise separately. Furthermore, by evaluating three parameters of wavelet coefficients' probability model using maximum likelihood method, self adaptively of the algorithm is achieved. The result of the detecting edge of contaminated aerial image by this method is satisfactory .