基于蒙特卡罗方法的边界检测不确定度估计  

Monte Carlo Simulation Based Uncertainty Evaluation of Edge Detection

作　　者：陈杰春[1] 张恒[1] 赵丽萍[2] 

机构地区：[1]东北电力大学自动化工程学院,吉林吉林132012 [2]东北电力大学理学院,吉林吉林132012

出　　处：《东北电力大学学报》2012年第3期75-78,共4页Journal of Northeast Electric Power University

基　　金：东北电力大学博士科研启动基金项目(BSJXM-200920)

摘　　要：探讨基于蒙特卡罗方法的边界检测不确定度评定方法。首先给出一种用于估计像素强度不确定度的方法,使用该方法,由一帧测量图像就可以估计出图像中任意一个像素的强度不确定度。在此基础上,将蒙特卡罗方法应用于边界特征检测不确定度的评定中。通过实验对文中给出的方法做了验证,实验结果表明,使用该方法可以准确估计出边界特征检测结果的不确定度。This paper discusses an approach toevaluating the uncertainty.of edge detection using.Monte Carlo Simulation. Firstly, the paper presents a method for evaluating the uncertainty associated with the intensity of a pixel. Using this method, the uncertainty associated with the intensity of a pixel can be evaluated using only one image. Furthermore, the Monte Carlo Simulation is used to evaluate the uncertainty associated with edge detection. The methods presented in this paper have been validated, and the exnerimental results show that the uncertainty associated with edge detection can be evaluated accurately using these methods.

关 键 词：边界检测 不确定度 评定 蒙特卡罗方法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]