首页 | 官方网站   微博 | 高级检索  
     

基于图像特征的边缘检测
引用本文:文婷,周激流,何坤. 基于图像特征的边缘检测[J]. 计算机工程与应用, 2011, 47(12): 189-191. DOI: 10.3778/j.issn.1002-8331.2011.12.053
作者姓名:文婷  周激流  何坤
作者单位:1.四川大学 电子信息学院,成都 610065 2.四川大学 计算机学院,成都 610065
摘    要:针对传统边缘检测算子在检测效果与抗噪方面的不足,提出基于图像特征的边缘检测方法,综合考虑图像的梯度特征、相位特征以及噪声的影响,以方向能量和由直方图差分算子计算的亮度梯度为图像特征进行边缘检测。对提取的特征值进行训练,得到边缘点与非边缘点的特征均值;依据近邻准则对检测图像的特征值分类,提取边缘;建立评价函数评价边缘的质量。仿真结果表明该算法可以取得很好的检测效果,同时具有一定的抗噪声性。

关 键 词:边缘检测  图像特征  方向能量  亮度梯度  
修稿时间: 

Edge detection based on image feature
WEN Ting,ZHOU Jiliu,HE Kun. Edge detection based on image feature[J]. Computer Engineering and Applications, 2011, 47(12): 189-191. DOI: 10.3778/j.issn.1002-8331.2011.12.053
Authors:WEN Ting  ZHOU Jiliu  HE Kun
Affiliation:1.College of Electronic and Information,Sichuan University,Chengdu 610065,China 2.College of Computer,Sichuan University,Chengdu 610065,China
Abstract:Accroding to the limitation of the traditional methods in detection quality and noise resisting ability,tlds paper presents a novel method based on image feature to detect edge.This method considers the effect of gradient feature,phase feature and noise at the same time.It makes oriented energy and brightness gradient got by histogram different operator as image fea- ture to detect edge.The computed feature values are trained and the feature mean values are got With the rule of distance closest, the pixels are classified and thereby the edge of the image is found.A function is established to evaluate the quality of edge.It is proved that the method can get good detection results and avoid the interference of noise to some extent.
Keywords:edge detection  image feature  oriented energy  brightness gradient
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《计算机工程与应用》浏览原始摘要信息
点击此处可从《计算机工程与应用》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号