Research On Network Monitoring System Based On Visual Tracking

In recent years,with the rapid development of machine vision,it plays an important role in the field of intelligent information.It covers a wide range of fields,including image processing,optical electronic technology,computer technology,as well as pattern recognition,artificial intelligence and so on.In the field of machine vision,it has a very important branch-visual tracking technology.The application of visual tracking technology can solve many problems in the fields of intelligent navigation,telemedicine,video monitoring and so on.This thesis studies the design and implementation of the network monitoring system based on visual tracking with video monitoring as the actual background.By analyzing and comparing the advantages and disadvantages of tracking algorithm based on detection,tracking algorithm based on Kalman Filter,tracking algorithm based on Particle Filter and tracking algorithm based on Meanshift,this thesis focuses on the research and design of improved Camshift tracking algorithm.First of all,this thesis summarizes the research status of related technologies at home and abroad.A practical application monitoring scheme is proposed,and then the overall scheme of the network monitoring system based on visual tracking is given according to the actual needs of the subject.This system adopts FPGA architecture,divides the system into five parts: image acquisition module,data cache module,core processing module,network transmission module and video display module,and gives the overall design framework of the system.Secondly,the image preprocessing technology is analyzed theoretically and verified by experiments.Research on mobile target detection algorithm,such as background difference method,frame difference method and Lucas Kanade optical flow method,analyze the detection principle,and verify the detection effect by experiments;And theoretically deduce the tracking algorithm based onfiltering and Mean Shift tracking algorithm,and analyze the effect of the above tracking algorithm by experiments.In order to solve the problems of occlusion and similar color in matching based tracking algorithm,an improved Camshift tracking algorithm is designed.By adding surf operator and fusion extended Kalman filter,the robustness and accuracy of target tracking are improved,and the performance of tracking algorithm is improved.Finally,the hardware structure and software design of this system are given,including the selection of components in each part and the schematic design of the whole system.The dynamic fast Medianfilter algorithm is used to preprocess the image.At the same time,Verilog HDL hardware description language is used to complete the logic design of each module,and the simulation experiment is used to verify.After the debugging of software and hardware,the performance test of the system is carried out.The experiment proves that the tracking target of the system is little affected by occlusion and similar color,and it can achieve the characteristics of effective image acquisition and remote data transmission.