Object tracking based on the improved particle filter method using on the bionic eye PTZ

Our bionic eye PTZ requires the tracking target at the central field of view of the camera, which means it is so important to realize the target tracking well in the first step. The particle filter method is famous for its robust tracking performance in cluttered environments. However, most methods are in the mode of moving object and stationary camera and they are not utilizing so well on the bionic eye PTZ since the camera in our project needs real-time motion. In this paper, we proposed an improved particle filter based on the SKL (Symmetric Kullback-Leibler divergence) similarity measure to realize object tracking and a closed-loop control model based on speed regulation to keep the target at the centre of the camera. The experiment results show that our system can track the moving object well and can always keep the object in the middle of the field of the view.