摘要
针对问题建立了运动模型,应用Kalm an滤波完成运动预测,在关联匹配中提出了一种代价函数,并通过分析运动特性提出了几种补偿算法,有效解决了跟踪中的重合、目标暂时消失等问题,实现了对目标的正确跟踪。
It establishes motion model to describe the problem. Then it uses Kalman filter to forecast the tracks. After that, a cost formula is given for data association. To solve the problems of superposition and disappearing, it brings forward the compensating algorithms based on analyzing the motion traits.
出处
《计算机与数字工程》
2006年第9期140-143,共4页
Computer & Digital Engineering
关键词
图像序列
Kslman滤波
关联匹配
补偿算法
Image Sequences, Kalman Filter, Data Association, Compensating Algorithms
