Research And Realization Of The Target Detection And Tracking System With Infrared Video Image Sequence On

Infrared guidance technology is widely used on reconnaissance,surveillance and guidance in the military field.What's more,Infrared guidance develops rapidly in medical,civil and other fields.The infrared target detection and tracking technology is an core content of infrared guidance technology.For those years,many experts and scholars have been working on the technology,and made a lot of detection and tracking algorithms.Based on the life of the common goal of the human body and car infrared characteristics,infrared video image sequence target detection and tracking technology is studied under Lab VIEW platform.The main contents are as follows:Firstiy,this paper analyzes the principle of inter-frame difference and background subtraction on the experiments of the human body and car.Experimental results show that the temperature difference of the human target and the surrounding background is large,so it is easy to extract target from the background.But the gray value of car is similar to the gray value of its background.The test results of inter-frame difference method and background subtraction are problematic.Therefore,in order to cancel the problems that the target value is similar with the gray background,background suppression method is proposed to reduce background interference and highlight target.By testing Background suppression method,indicating inter-frame difference method and background subtraction method,the target can be extracted well from a similar background in gray value.Secondly,the using of Meanshift algorithm in target tracking is analyzed simply.Furthermore,this paper does experiments about its change of scale and target occlusion.From the experimental results,the classical mean shift method is blocked to track the target whose size is changed and sheltered a large of area.As a consequence,the outline of the window function to be updated is proposed in this paper to handle the problem of size change.In order to solve the problem of tracking failure,Kalman filter is used to estimate the next time motioning of target.Then reuse the Meanshift algorithm to track target.The improved algorithm for tracking experiment results show that the improved tracking algorithm can solve the track the target that changes size and is obscured problems,And furthermore,the improved tracking algorithm ensure the tracking accuracy.Finally,using the LabVIEW to build the target detection and tracking system experiment platform of infrared video sequence.Each of the major functional modules are introduced detailedly,and given the results of the analysis.The system interface using LabVIEW allows users to operate more conveniently and easily to implement and maintain.Using IMAQ Vision and MATLAB to develop the system federatively can make the implementation of the algorithm more conveniently and the development cycle of the system shorter.