基于卡曼滤波与均值偏移算法的目标跟踪

针对变化场景下的目标鲁棒跟踪,提出了一种结合均值漂移与Kalman滤波的跟踪算法.利用YCbCr特征空间进行目标描述,使用Kalman滤波对目标运动速度和空间位置进行预测.根据干扰的不同情况,使用不同的比例因子将两算法的跟踪结果线性加权得到目标的最终位置,并利用一种比较科学的模型更新策略,减轻了模型漂移的影响,视频序列跟踪结果表明,提出的方法能够稳定地进行跟踪.

Target tracking based on Kalman filter and mean-shift

To deal with the robustness of object tracking in the time-variant scene,an algorithm combined Kalman filter with Mean-Shift algorithm is proposed in this paper.We employ YCbCr features to describe the object,and Kalman filter is used to predict the position and velocity of the target.According to different disturbance circumstances,the two algorithms tracking results are done with liner weight method by using different scale factors to get the final position of the target.Furthermore,a model update strategy is utilized to alleviate the model drift.Experimental results show the good performances of the proposed algorithm.