一种新的模糊图像边缘检测方法

边缘检测技术是图像预处理中最重要和最困难的任务之一，本提出了一种新的模糊图像边缘检测算法，算法中图像所对应的模糊特征平面通过一个基于阈值的隶属函数来提取，在模糊特征平面上应用模糊增强算子对低灰度区域的大部分象素进行衰减运算，对高灰度区域的大部分象素进行增强运算来提高两个区域之间的对比度，图像的边缘采用min或max算子来提取，仿真结果表明，该算法是一种有效的边缘检测方法。

A Fuzzy Algorithm for Better Edge Detection

Abstract: Improving edge detection is much to be desired in image preprocessing. We improve edge detection with a novel algorithm using fuzzy sets to strengthen most elements in the high gray-level region and at the same time, weaken most elements in the low gray-level region. We extract the fuzzy characteristic plane corresponding to the image from the spatial domain using a new membership function based on the image threshold. A fuzzy operator operating on the fuzzy characteristic plane accomplishes strengthening of the high gray-level region and also weakening of the low gray-level region. We use max or min operator to obtain final edges of the image. Figs. 4, 5 and 6 give simulation results of the original image in Fig. 3. Fig. 5 shows the edge detection obatined with our novel algorithm. Fig. 4 shows the edge detection obtained without strengthening of high gray-level region and weakening of the low gray-level region. Fig. 6 shows the edge detection obtained with the well-known method of Pal and King[4].