变电站外围监视入侵鼠类的目标跟踪研究  

Research on Target Tracking of Monitoring Rats Invasion Around Substation

作　　者：汤健飞 卓旭升[1] 祁星辰 王为 TANG Jian-fei;ZHUO Xu-sheng;QI Xing-chen;WANG Wei(School of Electrical and Information Engineering,Wuhan Institute of Technology,Wuhan 430205,China)

机构地区：[1]武汉工程大学电气信息学院,武汉430205

出　　处：《自动化与仪表》2022年第1期60-65,80,共7页Automation & Instrumentation

基　　金：湖北省教育厅科学技术研究项目(B2015317)。

摘　　要：针对变电站围墙外鼠类动物的监视问题,提出一种基于监控视频图像处理的多特征融合Camshift目标跟踪算法。采用多特征融合的方式,消除背景干扰,并引入线性外推法预测机制,通过预测目标的空间位置信息,矫正Camshift算法搜索框的中心位置,解决因目标受到遮挡或运动速度过快,导致跟踪框漂移的问题。与传统的Camshift算法进行对比,实验结果表明,该算法能有效消除背景干扰,且鼠类目标被严重遮挡或快速运动时,仍能准确跟踪。Aiming at the problem of real-time tracking of rats outside the substation,a Camshift tracking algorithm of multi-feature fusion based on surveillance video processing was proposed. Multi-feature fusion is used to eliminate background interference. In order to solve the problem that the tracking box drift due to the target being occluded or moving too fast,introduce a linear extrapolation prediction mechanism to correct the search box of the Camshift algorithm by prediction the spatial position information of the target. The results of a comparison experiment with the traditional Camshift algorithm show that this algorithm can effectively eliminate background interference and rats can still be accurately tracked when they are severely occluded or moving fast.

关 键 词：目标跟踪 CAMSHIFT算法 多特征融合 线性外推法 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]