一种优化的运动目标定位与跟踪算法研究  

An Optimization of Movement Target Location and Tracking Algorithm Research

作　　者：张晓春[1] 孟令军[1] 韩朝辉[1] 

机构地区：[1]中北大学仪器科学与动态测试教育部重点实验室电子测试技术国家重点实验室,太原030051

出　　处：《计算机测量与控制》2015年第4期1266-1268,共3页Computer Measurement &Control

基　　金：国家自然科学基金(51275491);国家自然基金重大研究计划(91123036)

摘　　要：运动目标检测与跟踪技术可实现对运动目标的检测、定位和跟踪,具有重要的理论意义和实用价值,已经成为各国学者研究的热点方向之一;为了达到更准确和有效的运动目标定位与跟踪的目的,该算法针对目标检测与目标跟踪两个方面,创新点在于目标跟踪;目标检测采用帧间差分法和背景差分法对比择优的方法,目标跟踪采用重心跟踪与卡尔曼滤波跟踪相结合的方法;通过对该种算法进行试验,得出的结果为:该算法的实现效果超越了传统的目标定位与跟踪算法;试验结论:该跟踪算法超越了传统的矩心、重心和卡尔曼滤波跟踪算法的单独跟踪效果,而且运算较快,同时卡尔曼滤波算法的预测与检测性大大降低了错误率,有效地改进了传统目标定位与跟踪算法。Moving target detection and tracking technology, which have important theoretical significance and practical value and have become one of the branches which international scholars have studied as hot topic, can realize the detection, location and tracking of moving targets. In order to achieve the purpose that a more accurate and effective location and tracking the Moving Target, the algorithm mainly study two aspects that the target detection and target tracking, the innovation lies in the Target tracking. The Target Detection adopts the method comparing the interframe difference and background difference, the Target Tracking adopts the method combining the center of gravity Tracking with the Kalman Filter Tracking. By testing the algorithm, the result as follows： the implementation effect of the algorithm goes beyond the traditional targeting and tracking algorithm. The test conclusion as follows： this goes beyond the separate tracking performance of the traditional centroid, the center of gravity and the Kalman Filter tracking, and faster computing. Simultaneously, the prediction and de tection of the Kalman Filtering algorithm greatly reduces the error rate and effectively improves the traditional Target location and tracking algorithm.

关 键 词：目标检测与跟踪 帧间差分 背景差分 卡尔曼滤波 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]