摘要
针对生产线上动态工件的跟踪分拣问题,提出了一种Kalman预测目标和Mean-Shift搜索目标综合应用的跟踪算法,实现了对履带上工件的动态跟踪。该算法首先利用Kalman滤波估计出后续运动目标的位置、速度和匹配范围,然后使用基于HSV色彩空间融合的Mean-Shift算法进行小范围搜索和目标匹配,最后将Mean-Shift算法得到的目标位置作为下一帧Kalman滤波器的输入参数使得后续状态具有预测的能力,迭代执行,直至搜索到目标为止。实验证明,该算法能够有效解决动态工件的跟踪和定位问题。
Aiming at the tracking and sorting problem of dynamic workpiece on production line, this paper proposes a kind of tracking algorithm that synthetically applies Kalman algorithm to predict the target and Mean-Shift algorithm to search the target; the dynamic workpiece tracking on the pedrail is realized. Firstly, this method uses Kalman filtering to predict the position, velocity and matching range of subsequent moving target, then uses the Mean-Shift algorithm based on HSV color space fusion to perform small range searching and target matching; and finally the target position obtained using the Mean-Shift algorithm is taken as the input parameter of Kalman filter for the next frame, which makes the subsequent state have predictive capability. This process is executed iteratively until the target is found. Experimental results show that this algorithm can effectively solve the tracking and location problem of dynamic workpiece.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2012年第12期2796-2802,共7页
Chinese Journal of Scientific Instrument
基金
国家863计划(2012AA041405)
沈阳市工业科技攻关项目(F12-010-2-00)资助
