摘要
针对传统边缘检测算法忽视图像不确定性的问题,研究了基于D-S证据理论的边缘检测方法。首先介绍了传统的边缘检测方法(一阶微分算子、二阶微分算子,以及在二阶微分算子上改进而来的最优算子法),并结合仿真结果分析了传统的边缘检测算法的优缺点。传统的边缘检测方法利用微分算子求出梯度后,使用一个阈值进行决策,判断该像素是边缘与否,采用硬阈值划分的方法忽略了像素点的不确定性。基于此,研究并实现了基于证据理论的图像边缘检测的方法,并将电力图像检测领域与传统的边缘检测方法进行了对比分析。结果表明,基于证据理论的边缘检测算法具有很好抗噪声能力,能够较少地检测出伪边缘,同时又检测出了相对完整的边缘。
Aiming at the problem that the traditional edge detection algorithm ignores the image uncertainty,the edge detection method based on D-S evidence theory is studied in this paper.Firstly,this paper introduces the traditional edge detection methods,i.e.,first-order differential operator,second-order differential operator and the improved optimal operator on the second-order differential operator,and analyzes the advantages and disadvantages of the traditional edge detection algorithm based on the simulation results.The traditional edge detection method uses a threshold to decide whether the pixel is edge or not after the gradient is calculated by differential operator.The method of hard threshold partition ignores the uncertainty of Pixel points,in view of this problem,the method of image edge detection based on evidence theory is studied and implemented.Compared with the traditional edge detection methods,the results show that the edge detection algorithm based on the theory of evidence has good anti-noise ability and can detect less pseudo edges.And at the same time,it can detect relatively complete edges.
作者
杨旭
李君
宋美艳
董小兵
YANG Xu;LI Jun;SONG Mei-yan;DONG Xiao-bing(State Grid Integrated Energy Service Group Co.Ltd.,Beijing 100176,China;North China Electric Power University,Beijing 102206,China;Xi’an Jiaotong University,Xi’an 710049,China;State Grid Shaanxi Economic and Technological Research Institute,Xi’an 710075,China)
出处
《电力学报》
2020年第4期360-368,共9页
Journal of Electric Power
关键词
边缘检测
一阶微分算子
二阶微分算子
最优算子法
edge detection
order differential operator
second order differential operator
optimal operator method
作者简介
杨旭(1982-),男,硕士,国网(北京)综合能源服务有限公司副总经理,研究方向为电力需求侧管理;李君(1983-),女,助理工程师,研究方向为高电压与绝缘技术;宋美艳(1987-),女,硕士,从事机械设备状态检修方面的研究工作,songmeiyan@xjtu.edu.cn;董小兵(1981-),男,出版中级,从事电力行业期刊编辑出版方面的工作。
