基于冗余小波变换的运动目标检测算法  

Moving object detection algorithm based on redundant wavelet transform

作　　者：王建勋[1] 于明[1] 夏颖[1] 陈冀川[1] 

机构地区：[1]河北工业大学计算机科学与软件学院,天津300401

出　　处：《河北工业大学学报》2013年第2期20-23,共4页Journal of Hebei University of Technology

基　　金：河北省自然科学基金重点项目(ZD200911)

摘　　要：针对传统运动目标检测方法帧差法、光流法等提取运动目标的缺陷与不足,提出了一种新的基于冗余小波变换和DT网格的运动目标检测算法,即在冗余小波域提取特征点生成Delaunay三角形网格,并自适应获得潜在运动区域,从而获得运动目标区域实现运动目标检测.实验结果表明,本文提出的算法可以有效地提取运动目标,效果要好于传统的运动目标检测方法,对于室外交通监控和室内监控的运动目标检测同样适用,取得不错的效果.An object detection method which can be used in indoor and outdoor monitoring is proposed in this paper.Improving shortcomings and disadvantages of moving object extraction in traditional inter-frame difference method,it puts forward a moving object recognition algorithm based on redundant wavelet transform,namely,extraction of feature points in redundant wavelet domain to generate a delaunay triangulation grid and adaptive potential movement area,then moving object area.The experimental results show that the proposed algorithm can effectively extract moving object,which is better than traditional frame difference method.

关 键 词：冗余小波变换 目标检测 视频挖掘 DT网格 特征点 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]