改进的自适应形态学边缘检测方法  

An Improved Adaptive Morphological Edge Detection Method

作　　者：葛阿雷 史伟[1] 

机构地区：[1]宁夏大学数学计算机学院,宁夏银川750021

出　　处：《宁夏大学学报（自然科学版）》2016年第1期34-38,共5页Journal of Ningxia University(Natural Science Edition)

基　　金：国家社会科学基金西部项目(11XMZ033);宁夏自然科学基金资助项目(NZ1123)

摘　　要：针对传统边缘检测算法对含噪图像检测效果的局限性,给出了一种改进的自适应形态学边缘检测算法.该算法首先利用2种不同结构和尺寸的矩阵元素对图像进行滤波,然后用不同方向的结构元素对图像进行形态学边缘检测并根据图像边缘信息设置各方向的权值,最后根据不同的权值对各个方向检测到的边缘结果进行加权求和.实验表明,改进的形态学边缘检测算子能够更好地减弱噪声,在准确定位图像边缘的前提下,保留了更多的边缘细节信息.In the view of the limitations of the traditional edge detection algorithm for noisy image detection,an new method which improved pre-existing adaptive morphological edge detection algorithm is proposed.Firstly,the different structures,different sizes of structure elements are used to filter image.Then,the different directions of structure elements are adopted to detect image edge based on morphology and set the weight of directions through characteristics of the image edge information.Finally,the weighted sum of detected edge results are calculated by the different weights of each party.Experiment shows that the improved morphological edge detection operators can better reduce the noise and preserve the more edge details in the accurate positioning of image edge.

关 键 词：形态学边缘检测 含噪图像 多结构 加权求和 

分 类 号：O236[理学—运筹学与控制论] TP391.9[理学—数学]