Abstract Oil spill detection is one of the very important application fields in Synthetic Aperture Radar (SAR) images. Wavelet method is an edge detection method broadly researched in resent years. Normal oil spill detection uses the first and second derivatives of a Gaussian function as the basic wavelet function, which has slow computing speed. This paper utilizes zero-antisymmetrical dyadic wavelet, using multi-resolution analysis and multi-scale integration to get oil spill edge images. The results demonstrate that it is a useful and promising method for oil spill detection in SAR images.