结合Wiener滤波和形态学的图像边缘检测方法  

Method of Image Edge Detection Based on Wiener Filter and Morphology

作　　者：黄剑玲[1] 郑雪梅[2] 

机构地区：[1]上饶师范学院数学与计算机系,江西上饶334000 [2]上饶师范学院物理系,江西上饶334000

出　　处：《郑州大学学报（理学版）》2010年第2期57-60,共4页Journal of Zhengzhou University:Natural Science Edition

摘　　要：提出了一种将Wiener滤波和数学形态学相结合的图像边缘检测方法.首先用自适应Wiener滤波对含噪图像去噪,然后构造6种具有代表性的结构元素,用改进的形态学边缘检测算子对图像进行边缘检测,自适应确定权重,将多结构元的边缘检测信息加权求和,得到在噪声存在条件下较为理想的图像边缘.实验结果表明,该算法能够有效地抑制噪声,检测的边缘较清晰、连续,其检测效果优于传统边缘检测算法.A new image edge detection method that combines Wiener filter and mathematical morphology is presented. Firstly, image noise is eliminated with an adaptive Wiener filter . Then, image edge is detected by using six representative structural elements and improved morphological edge detection operator. The edges of various sizes are combined according to adaptive weights. Finally, the more ideal image edge under the environment of existing noise is obtained. The experimental result shows that algorithm can suppress noise effectively, detect clearer and continuous edge. Detection results are superior over the traditional edge detection algorithm.

关 键 词：WIENER滤波 形态学 边缘检测 结构元素 自适应 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]