一种自适应特征选择的运动目标实时跟踪算法

An Adaptive Feature Selecting Real-time Moving Object Tracking Algorithm

提出自适应特征选择算法,利用背景信息及目标信息建立特征分类器,并在跟踪过程中不断更新特征分类器;提出采用光流算法对运动区域进行粗预测,然后利用特征分类器及meanshift算法对目标进行跟踪.实验结果表明,该算法可以根据不同的背景信息自适应的选择特征,对于跟踪过程中存在形变、遮挡以及背景出现干扰或光照变化等情况,依然可以对目标进行稳定的实时跟踪.

The paper proposed an adaptive features selecting algorithm. Features classifiers are constructed utilizing object information and background information, and updated during tracking. Optic-flow model is used to predict motion area roughly, and then object is tracked by utilizing clssifiers and Meanshift algorithm. The experiment result shows that the tracking algorithm we proposed can adaptively select features for tracking utilizing different background information, in the existence of covering, appearance changed, clutter in the background and illumination changing, we can still track objects stably and in realtime.