融合提升小波和形态学熵权的医学图像边缘检测

MEDICAL IMAGE EDGE DETECTION BY FUSING LIFTING WAVELET AND MORPHOLOGY ENTROPY WEIGHT

为了提高医学图像边缘检测定位的精度,结合基于小波变换和数学形态学边缘检测算法的优点,提出一种融合提升小波和多尺度形态学熵权边缘检测算法。首先应用提升小波边缘检测算法提取边缘,再由多尺度形态学算子进行边缘检测,依据各尺度下边缘图像的信息熵确定权值进而求和得到边缘图像,最后对两种方法得到的边缘图像进行融合。实验结果表明,与传统边缘检测算法相比,该算法融合规则简单,边缘精度高并且定位准确,是一种有效的图像边缘检测算法。

In order to improve positioning accuracy of medical image edge detection,we propose an edge detection algorithm which fuses the lifting wavelet and multi-scale morphology entropy weight based on the combination of the advantages of wavelet transform and mathematical morphology-based edge detection algorithm.First,the edge detection method based on lifting wavelet is applied to extract the edges.Then the multi-scale morphological operators are used to detect the edges,according to the information entropies of edge images in various scales the weights are determined,and the edge images are further derived by summing operation.Finally,the edge images obtained from the above two methods are fused.Experimental results show that,compared with traditional edge detection algorithm,the proposed algorithm has simple fusion rule,high edge precision and accurate positioning.It is an effective image edge detection algorithm.