基于二级特征匹配的卡尔曼滤波跟踪算法  

Object Tracking Algorithm based on Kalman Filter and Two Level Feature Matching

作　　者：杨平[1] 厉小润[1] 

机构地区：[1]浙江大学电气工程学院,浙江杭州310027

出　　处：《有色冶金设计与研究》2011年第4期34-37,共4页Nonferrous Metals Engineering & Research

基　　金：浙江省重大科技专项重点工业项目(2010C11069)

摘　　要：运动目标的检测和跟踪是计算机视觉领域的研究热点和难点。本文针对复杂的交通应用背景,提出了基于二级特征匹配的卡尔曼滤波跟踪算法。首先在背景建模中引入一个差异深度矩阵以抑制噪声影响,利用基于背景差分的目标检测确定目标区域,然后运用卡尔曼预测目标位置,并以目标区域的几何特征和核直方图特征对当前帧检测到的目标模板和预测区域内的目标进行二级特征匹配。实验表明,该算法能够在保证多运动目标跟踪的准确有效性的基础上达到实时性应用要求。The moving object detection and tracking is a hot and difficult spot in computer vision field. In order to make application for complex background, an object tracking algorithm based on kalman filter and two level feature matching is presented in this paper. First, a depth matrix is introduced in the background modeling to reduce the noise influence and then use Kalman filter to predict the object position, and make the two level feature matching based on geometrical characteristics and kernel histogram. The experiment shows that this algorithm can meet the requirements of real-time and validity.

关 键 词：背景差分 目标检测 目标跟踪 卡尔曼滤波 二级特征匹配 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]