A New Kind of Image Edge Detection Based on The Theory of The Adaptive Lifting Wavelet and Morphology

This article describes how in traditional edge detection it is prone to defects such as fuzzy positioning, and noise influence. This article proposes a type of edge detection algorithm which combines lifting wavelet transform and adaptive mathematical morphology, which makes a lifting wavelet to analyze the wood cell image. Then, the high-frequency part is detected by using the algorithm fusing the wavelet packet and the rapid-combining multi-scale wavelet, which controls noise effectively; while for the low frequency part is detected with modified adaptive mathematical morphology, to locate the exact details. The final result will processes the edge of the image using "algebra" algorithm fusion. The example for a wood cell image which illustrates the algorithm is to detect the cell boundary relatively clearly, and effectively suppress the noise.