噪声图像边缘检测方法的研究  

Study of edge detection method for noisy image

作　　者：徐艳蕾[1,2] 赵继印[1,3] 焦玉斌[4] 

机构地区：[1]吉林大学通信工程学院,长春130022 [2]吉林农业大学信息技术学院,长春130118 [3]大连民族学院机电信息工程学院,辽宁大连116600 [4]装甲兵技术学院控制系,长春130117

出　　处：《计算机应用研究》2009年第1期387-389,共3页Application Research of Computers

基　　金：大连民族学院人才引进科技基金资助项目(20086201)

摘　　要：对图像中常见的高斯噪声和椒盐噪声的特性进行了分析,对含有这两种噪声的数字图像的边缘检测方法进行了研究,尤其对基于顺序形态学的噪声图像边缘检测方法进行了深入的分析研究,基于顺序形态学理论提出了针对分别含有高斯噪声和椒盐噪声的图像边缘检测方法。通过仿真验证表明,对于含有以上两种噪声的图像,提出的方法不仅能够去除噪声而且能够提取出准确清晰的边缘。The image edge detection is an important tool of image processing and also is the foundation of pattern recognition and computer vision. This paper analyzed the specialities of Gauss noise and salt and pepper noise, studied the edge detection methods for images contained by these two noises. And researched very deeply especially to the edge detection method of noisy image based on the order morphology, proposed the corresponding edge detection method for image contained respectively by Gauss noise and salt and pepper noise. At last, simulation experiment is done and the result shows that the method not only can eliminate the noise but can extract the clear and exact edge.

关 键 词：高斯噪声 椒盐噪声 噪声图像 边缘检测 顺序形态学 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]