视频图像序列中运动目标的提取与跟踪  

Extracting and Tracking the Moving Object in Video Sequences

作　　者：刘雪[1] 王华杰[1] 常发亮[1] 

机构地区：[1]山东大学控制科学与工程学院,山东济南250061

出　　处：《自动化技术与应用》2007年第1期102-103,128,共3页Techniques of Automation and Applications

摘　　要：提出一种摄像机静止条件下视频图像运动目标提取与跟踪的方法。本文首先提出了一种有效的阴影去除算法检测运动目标,然后采用基于目标颜色直方图的相关匹配,使用Bhattacharyya系数度量目标模型与预测模型间的相似度,选出最优相似模型作为当前的目标模型,实现了对目标的跟踪。实验结果表明该算法是可行有效的。An algorithm of extracting and tracking moving targets in video sequences is proposed. First a shadow-removal scheme is designed to detect targets. Based on color histogram matching, Bhattacharyya coefficient is used to measure the similarity between template and possible target s model. The most similar model is chosen as the current target model.

关 键 词：跟踪 阴影去除 颜色直方图 BHATTACHARYYA系数 

分 类 号：TN919.81[电子电信—通信与信息系统]