针对目标跟踪过程中存在的光线变化及目标遮挡等复杂情况所导致的跟踪目标的丢失或跟踪错位等问题, 采用特征融合的策略, 将其用于粒子滤波框架中进行跟踪。利用特征融合加权直方图的方法来描述目标的颜色和纹理特征, 并对变化的目标自适应地更新参考目标模型,能有效地处理视频中由于旋转、遮挡、光线变化带来的影响。实验证明, 提出的算法在保证跟踪精确性的情况下, 具有较好的适应性及鲁棒性。

For the issues caused by the complexity of light changing and object occlusion in object tracking, multi-feature fusion strategy is used in particle filter tracking. The multi-feature weighted histogram is proposed, which can describe the color and texture characteristics of the tracking target, and updates the reference target model adaptively to the changed target. The algorithm can effectively deal with the impact of the rotation, the shutter, the light moves in the video. Experimental results show the proposed method is adaptability and robustness in the complex environment with high precision.