Research On Moving Target Detection And Tracking Based On Vector Flow In Guided Projectiles

Under the background of modernized and informational partial warfare,the demand for precision-guided weapons is getting higher and higher,that is to say the guided weapons should have precise strike capability against uncooperative mobile targets on the battlefield,however,it has certain limitations in distance and could no longer meet the needs of a moderm battlefield that the conventional tracking method of manually selecting targets through the seeker's returning screen.Meanwhile,conventional guided weapons seeker claims capable of resisting high overload,small size and low cost.Therefore,it is of great significance in improving the precision strike capability of man-made guided weapons to study the ground target autonomous detection and tracking on guided missile based on strapdown image seeker in order to realize the self-detection and target tracking with strapdown seeker.For this issue,the following works are done:Taking the guided projectiles that rotate at a low speed attack on the ground moving target as the background,the working principle and key theories of strapdown seeker is studied in detail,and the electronic image stabilization methods are designed and simulated aimed at the problem of image instability caused by the coupling of strapdown image seeker and projectile attitude as well as the performance evaluation index of the image stabilization is given.A moving object detection method based on improved optical flow method is proposed on the basis of video image stabilization.The moving target detection process of the method based on optical flow is analyzed in detail as well as the simulations under different image background conditions is done,and the advantages is found out between the algorithm in this paper with the others.Aiming at the issue of moving target real-time tracking in complicated background,the key problems in the process of moving target tracking are summarized,and the existing tracking algorithm based on template matching is improved.On this basis,a tracking algorithm based on kemelized correlation filters is designed,and the tracking algorithm is simulated for different moving targets in different scenarios.The advantage of the tracking algorithm based on kemelized correlation filters and the improved template matching tracking method are compared and analyzed.The effectiveness of the algorithm for continuous tracking of moving targets under complex ground conditions is verified.An experiment is designed according to the testing and tracking algorithm of moving target in paper,that is the process of man-made missile hitting moving targets is simulated by UAV flying seeker.The hardware platform and the experimental scheme are introduced,what's more,the configuration of the simulation software environment and the specific software architecture are designed,and the value and feasibility in engineering of this detection and tracking algorithm is certified by the experiment results of the simulations.