A method of multi-scale edge detection based on lifting scheme and fusion rule

Recently, the edge detection methods based on wavelet transform are mainly to extract image edge of high-frequency parts and to ignore some edge information of low-frequency ones. And the traditional wavelet transform is time and memory consuming. Therefore a method of multi-scale edge detection based on lifting scheme and fusion rule is proposed in this paper. This algorithm should be carried out first to decompose the source image by lifting scheme. Then the edges of high-frequency and low-frequency sub-images are extracted separately. The final edge image is obtained through the fusion rule of average energy. Experimental results showed that the algorithm has advantages of quick operation, efficient suppression of noise and high detection accuracy.