Objective Performance Evaluation of Infrared Moving Foreground Segmentation

摘要：

　　Moving foreground segmentation has been widely used for many infrared video applications.In this work,a metric is presented to evaluate the performance of infrared moving foreground segmentation objectively.This metric is constructed on the basis of the criterion in Joint Segmentation (JSEG).Difference image is the result of consecutive frame subtraction and the pixels of it are classified into two classes:changed pixels and fixed pixels.With making a comparison between the distributions of classes in segmented foreground regions and background regions,it can accurately reflect the difference between segmented foreground and background.If the difference is greater,moving foreground segmentation is better.Experiments are performed on variety situation of moving foreground segmentation,showing that the proposed metric is effective.

关键词： Performance Evaluation Foreground Segmentation Infrared Video

作者: Chaobo Mina Junju Zhanga Benkang Changa

作者单位: School of Electronic Engineering and Optic-electronic Technology,Nanjing University of Science and Technology,Nanjing,Jiangsu Province,P.R.China,210094

会议类型: 国内会议

会议名称: 2012年全国光学工程博士生论坛

会议地点: 南京

会议语种:英文

页码: 1-7

在线出版日期: 2012-06-08（万方平台首次上网日期，不代表论文的发表时间）