Footprint Edge Detection Algorithm Based on Fusion Technology

Abstract

A novel algorithm of footprint edge detection is proposed. It combines wavelet transform with mathematical morphology and takes the full advantage of the smooth filter wave of Gauss function, and uses Gauss weighted fusion strategies to determine image edge position. In wavelet domain, the edges of high frequency sub-image are detected by using a wavelet modulus maximum method, while the edges of low frequency sub-image are detected by using a mathematical morphology method. Finally the fusion rules of Gauss function weighted is used to fuse sub-image edges of high frequency and low frequency. Experimental results show that this algorithm is possessed of the ability of strong generalization, effective controlling noise and overcoming the distortion of image edges. It can reappear the image edge information better, so it is an effective algorithm of the image edge detection.