Mean Shift Particle Filter Video Target Tracking Algorithm Based On Color And Depth Features

Video-based moving target tracking has always been a research hotspot in the field of computer vision.It combines image processing,pattern recognition,and computer application related advanced technologies,and has been more and more widely used in public safety,military guidance,and traffic monitoring.The performance of the moving target tracking algorithm has a direct impact on the tracking effect.Through continuous exploration of the field,researchers have proposed many target tracking algorithms,but in the face of complex tracking environments,the effect is not satisfactory.This paper focuses on the problems of low algorithm tracking efficiency and poor robustness in complex backgrounds,and improves the video target tracking algorithm.Main tasks as follows:(1)In view of the single feature of the traditional particle filtering algorithm based on color features,this paper proposes a high-dimensional feature model that combines color features and depth features.The color features extract the surface color features of the target,while the depth features extract the target's spatial information,the combination of the two features more accurately represents the target's information,and solves the problem of the uniqueness of the target's characteristics.Experiments show that when faced with complex tracking environments such as illumination changes and occlusion,it effectively improves the robustness of the particle filter tracking algorithm.(2)By analyzing the traditional Mean Shift algorithm and particle filter algorithm,the tracking framework based on the combination of Mean Shift and particle filter algorithm is determined.The particle filter uses the Mean Shift method to continuously approximate the posterior density function,and uses the gradient information to regenerate the particles,so that more particles are located in the high probability distribution area of the posterior density,thereby reducing the number of particles and the calculation complexity,and effectively improving the particles.The tracking efficiency of the filtering algorithm improves the real-time tracking.(3)A Mean Shift particle filtering video target tracking algorithm based on color and depth features is proposed,which uses a combination of color features and depth features to more effectively locate the location of dynamic targets,and then uses the clustering effect of Mean Shift to improve particles Sampling efficiency.Through simulation experiments on the target tracking algorithm,we can see that the algorithm has a certain improved the tracking accuracy and tracking efficiency compared with the traditional particle filter and Mean Shift algorithm.The tracking method proposed in this paper not only ensures the tracking accuracy under complex backgrounds such as illumination changes,occlusions,fast motion and scale changes,but also reduces the calculation amount of the particle filter algorithm.Compared with the tracking algorithm,the method in this paper is applicable to a wider range of scenarios and is a tracking algorithm that takes into account both accuracy and efficiency.