摘要
传统的Camshift算法在跟踪的目标发生遮挡时容易导致跟踪失败,针对这种情况提出将Kalman滤波与Camshift相融合的算法。在运动的跟踪目标出现遮挡的情况下,根据Camshift算法得到的运动目标在上一帧的运动参数,利用卡尔曼滤波对当前帧运动目标的参数进行预测,从而保证了实时跟踪。实验对动态遮挡和静态遮挡两种情况分别进行了验证,均得到良好的效果。实验结果表明该算法具有比较好的实时性,而且能够有效的解决遮挡情况下的运动目标的跟踪等问题。
A new algorithm combining Kalman filtering and Camshift was proposed by to slove the poor tracking ability problem in occlusions.When moving target is in occlussions,in order to ensure real-time tracking,we use Kalman filtering to predict the current frame of the moving target parameters,based on the last parameters of the moving target which was got from Camshift algorithm.We have verified the dynamic and static shielding block respectively,and obtained good performances.The experimental results show that the improved algorithm can better the real-time property and the cover problem distinctly.
出处
《电子测量技术》
2011年第5期8-11,共4页
Electronic Measurement Technology
