基于时空上下文(SpatialTemporal Context, STC)的跟踪算法与大部分传统算法相比, 在实时性方面具有明显的优势。通过实验发现, STC算法存在由变形和遮挡引起的跟踪精度下降问题。针对该问题, 提出了一种改进方法, 该方法在原STC算法的基础上引入局部二值模式(Local Binary Pattern, LBP)和遮挡检测机制, 利用LBP特征来代替灰度特征, 当跟踪器检测出目标发生遮挡时, 停止分类器参数的更新。对于满足线性运动的目标, 利用卡尔曼滤波器对其进行位置预测以解决目标发生遮挡后的定位问题。实验证明, 所提出的改进算法能有效提升目标跟踪精度, 针对遮挡情况下的目标也展现出较高的鲁棒性。

Compared with most traditional tracking methods, the method of tracking via spatialtemporal context (STC) has significant advantage in real time performance. Extensive experimental results show that STC presents degradation of tracking accuracy caused by deformation and occlusion. To address this problem, an improved method adopting occlusiondetection strategy and local binary pattern (LBP) to replace the gray feature was proposed. When the occlusion is detected by the tracker, the updates of classifier’s parameters are stopped. For the object of linear motion, the Kalman filter was adopted to predict its location, which will address the problem when the object is occluded. Experimental results indicate that the proposed method improves the tracking accuracy, and it shows high robustness for the occluded objects.