基于改进的codebook算法的运动目标检测  

Detection of moving targets based on the improved codebook algorithm

作　　者：赵玉吉[1] 骆且 杜宇人[1] 

机构地区：[1]扬州大学信息工程学院,江苏扬州225127

出　　处：《扬州大学学报（自然科学版）》2015年第4期63-67,共5页Journal of Yangzhou University：Natural Science Edition

基　　金：国家自然科学基金资助项目(51273172)

摘　　要：针对传统的codebook算法在面临复杂的环境背景干扰下进行运动目标检测中存在算法冗余的问题,提出一种改进的codebook运动目标检测算法.通过改进的码本描绘背景中感兴趣的码元进行建模,同时选择YUV空间模型替换传统的RGB模型,并在目标检测阶段融入训练元素来解决背景像素变化的问题,以提高运动目标检测的计算效率和可靠性.仿真结果表明,该方法比传统的目标检测算法具有更高的实时性、鲁棒性和准确性.In the detection of moving targets, using codebook algorithm can effectively solve the interference problem of complex environmental background. Improved codebook algorithm is proposed to solve the algorithm redundancy problem of traditional codebook algorithm and improve algorithm performance. By improving codebook, it achieves the object of interest in the context of modeling symbols. In order to improve the efficiency and reliability of moving object detection, this paper uses the improved codebook algorithm to describe the interesting elements in the background for modeling, and chooses YUV space model to replace the traditional RGB model at the same time. Meanwhile the training element is integrated into the target detection phase to solve the problem of background pixels changes. The experiment results prove that this method has higher real-time performance, robustness and accuracy than the traditional methods of target detection.

关 键 词：运动目标检测 时间序列 codebook算法 YUV空间 训练元素 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]