多特征融合与Kalman滤波的CAMShift跟踪算法

CAMShift Tracking Algorithm Based on Multi-Feature Fusion and Kalman Filter

针对CAMShift算法在实际应用场景中受颜色和遮挡时跟踪失败的问题,提出一种多特征融合与Kalman滤波的CAMShift目标跟踪算法。多特征融合是在CAMShift算法基础上将边缘、纹理与颜色特征融合在一起,采用改进的Canny算子描述边缘特征,采用统一模式下的N-LBP构造纹理特征,并利用巴氏(Bhattacharyya)系数计算各个特征的自适应融合权值,通过不同特征之间的优势互补,增强特征的表达能力。当跟踪目标无遮挡时,使用CAMShift算法计算目标位置并更新Kalman滤波器参数,有遮挡时,使用Kalman滤波预测当前目标的位置,最后仿真实验表明,本文算法受环境影响小,相比CAMShift算法跟踪误差显著降低。

A CAMShift target tracking algorithm based on multi-feature fusion and Kalman filtering is proposed to solve the problem of tracking failure when CAMShift algorithm is blocked by color and occlusion in practical application scenarios.Multi-feature fusion is to fuse edge,texture and color features together based on CAMShift algorithm.The improved Canny operator is used to describe the edge feature,and the N-LBP in the unified mode is used to construct the texture feature.The Bhattacharyya coefficient is used to calculate the adaptive fusion weight of each feature.Through the complementary advantages between different features,the expression ability of the feature is enhanced.When there is no occlusion,CAMShift algorithm is used to calculate the target position and update the Kalman filter parameters.When there is occlusion,Kalman filter is used to predict the current target position.The simulation results show that the algorithm is less affected by the environment and the tracking error is significantly reduced compared with CAMShift algorithm.