首页 | 本学科首页   官方微博 | 高级检索  
     

基于彩色图像边缘差分的运动目标检测算法
引用本文:宋宁,尚振宏,刘辉,舒国锋.基于彩色图像边缘差分的运动目标检测算法[J].微型机与应用,2011,30(24):36-38,42.
作者姓名:宋宁  尚振宏  刘辉  舒国锋
作者单位:1. 昆明理工大学信息工程与自动化学院,云南昆明,610000
2. 昆明理工大学云南省计算机技术应用重点实验室,云南昆明,610000
基金项目:基金项目:昆明市科技局重点项目“多PTZ摄像头融合跟踪技术在社会治安视频监控中的应用”
摘    要:针对目前常用的运动目标提取易受到噪声影响、易出现阴影和误检漏检等情况,提出了一种基于Sobel算子的彩色边缘图像检测和帧差分相结合的检测方法。首先用Sobel算子提取视频流中连续4帧图像的彩色边缘图像,然后将边缘图像进行隔帧差分相与,提取出较精确的运动目标边缘轮廓。提取的轮廓经过一系列的形态学操作填充,可得到完整的运动目标。实验结果表明,该方法对运动目标边缘轮廓提取准确,抗噪能力强,且鲁棒性好。

关 键 词:彩色边缘检测  隔帧差分  运动目标提取

Moving object detection algorithm based on color image edge-differencing extraction
Song Ning,Shang Zhenhong,Liu Hui,Shu Guofeng.Moving object detection algorithm based on color image edge-differencing extraction[J].Microcomputer & its Applications,2011,30(24):36-38,42.
Authors:Song Ning  Shang Zhenhong  Liu Hui  Shu Guofeng
Affiliation:Song Ning1,Shang Zhenhong1,Liu Hui2,Shu Guofeng2(1.School of Information Engineering and Automation Institute,Kunming University of Science and Technology,Kunming 610000,China,2.Computer Technology Application Key Laboratories of Yunnan,School of Kunming University of Science and Technology,China)
Abstract:To improve the moving object detection which susceptible to noises and shadows, which has incorrect leak detection, an algorithm based on color edge extraction of Sobel, kenal and frame differencing is presented in this paper. Firstly, four continuous frames of color edge image are extracted by the edge extraction from video. Then, the edge-image-differencing is made so as to extract edge contours of moving object. Finally, the edge contours are filled with morphology method. The test results shows that the proposed algorithm can detect the moving object accurately and quickly with less noise.
Keywords:color edge extraction  frame-differencing  moving object detection
本文献已被 CNKI 万方数据 等数据库收录!
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