Abstract: An edge detection approach combining Canny with wavelet transform is proposed. The source image is decomposed by wavelet and, on the different decomposition levels, the edges of low-frequency sub-images are detected using Canny operator, while the edges of high-frequency sub-images are detected using wavelet transform method. Then, the two edge images are fused into one complying with relevant rules to obtain an integrated and clear edge image. This edge detection method combines the advantages of both wavelet transform and Canny methods together to fuse the two edge information obtained by different methods, thus suppressing effectively the noises with the consecutive and clear edges kept up. Experimental results showed that the approach is superior to either Canny method or wavelet transform method used alone.