基于颜色与深度信息特征融合的一种多目标跟踪新算法

A novel multi-object tracking algorithm based on RJMCMC with RGB and depth information fusion

提出了一种基于颜色与深度信息特征融合的可逆跳的马尔科夫链蒙特长洛(RJMCMC)多目标跟踪算法。首先,融合彩色信息和深度信息对运动目标进行检测;然后,根据多目标检测的结果建立观测似然模型,并构建合理的状态转移模型;最后,通过RJMCMC粒子滤波算法实现多目标跟踪。实验结果表明,本文提出的多目标跟踪算法具有较强的鲁棒性,能够稳定的跟踪多目标,具有较高的准确率。

Multi-object tracking technique is one of the challenging issues in computer vision.In this paper,we propose a multi-object tracking algorithm based on RJMCMC with RGB and depth information fusion.Firstly,the moving objects are detected by fusing color information and depth information.Then,the observation likelihood model is established according to the result of object detection,and a reasonable state transition model is also established.Finally,the multi-object tracking result is obtained using RJMCMC particle filtering algorithm.Experimental results show that the proposed method has strong robustness.The algorithm can steadily track multiple targets and has higher accuracy.