Research On The Key Algorithm Of Vision Tracking Of Moving Subjects Under Blocking Environment

Machine vision is one of the primary means which uses computers to replace human eyes and brains,to gain information and to understand the world.Moving target tracking technology is an important research direction in machine vision,it has been widely applied in various areas such as image processing,artificial intelligence and many other directions in people's daily life.However,the environment around the tracked target are usually complicated,there are a variety of interference factors affect the tracking effect.Therefore,on the basis of deep studying of the theory and application fields of machine vision,this paper focuses on the tracking technology of moving objects in the occluded environment.Moving target detection research first briefly introduces the related image processing knowledge to be used in the target detection process,which guarantees a good image foundation for the follow-up detection and tracking operations.Then it focuses on the classical algorithm of moving target detection,detailing their principle,implementation process and their respective scenarios.Finally,through the experimental simulation,the detection effect of each method is inspected one by one.The research of moving target tracking in unshaded environment firstly introduces the tracking algorithms one by one according to the different tracking methods.Through the introduction of non-parametric density estimation.The basic principle and realization process of MeanShift tracking algorithm are introduced in detail.Then the MeanShift algorithm is applied to the field of moving target tracking,and the application steps of Mean Shift algorithm in the process of target tracking are explained in detail.Finally,through a set of comparative experiments,the tracking effect of the MeanShift tracking algorithm in the complex environment without occlusion and with occlusion can be shown,respectively.In the occlusion of the target tracking research,this part is the focus of this article,Firstly,the principle of Kalman algorithm and its application in the process of target tracking are introduced.Then it gives a detailed analysis and solution to the occlusion problem.Then the improvement of the traditional MeanShift tracking algorithm is focused.Firstly,the tracking region and partition method are redefined to make full use of the target space information.Then the Kalman prediction mechanism prediction is added to the new algorithm to achieve the reasonable prediction of the moving target trajectory in the occlusion case and to estimate the possible position of the moving target in the subsequent image sequence and then to complete the tracking of the moving target.Finally,a series of comparative experiments are conducted to verify that the modified MeanShift algorithm is satisfactory for tracking moving targets under occlusion.