Target Recognition Algorithm For Forward-looking Scanning Radar And Software Development

Target recognition is based on forward-looking images contains target position information extracted by analyzing the correlation and feature information between the front-view radar images and optical images in the same geographical location.The development of this research is conducive to improving the performance of autonomous navigation,positioning and reconnaissance capabilities of aircraft.It can also assist operators to analyze and judge the target area.So,the application value is high in the fields of airdrop,weapon guidance,and aircraft landing guidance.This thesis is focused on the radar super-resolution method and target recognition method.Meanwhile,the interpretation system is developed via transplanting the target recognition algorithm,which realizes the engineering realization of the target recognition method.The main research content is as follows:1.Aiming at the problem of low resolution of the radar images,the echo model of the forward-looking radar is analyzed,and one radar super-resolution method based on the maximum likelihood framework is studied.Under the premise of high efficiency,the resolution of radar images has been effectively improved.2.By studying the preprocessing method of the target images,which can effectively improve the accuracy of the subsequent target recognition method.3.A target recognition method based on morphological and spatial features is proposed,which can effectively distinguish the false regions in the image.So the target recognition in the presence of interference regions is achieved.4.The structure and implementation scheme of the client and the database of interpretation system are given.Then the engineering implementation of interpretation system is completed,through the algorithm migration of target recognition method.Meanwhile,aiming at the requirement of real-time and stability,improvement method based on multi-threading is given,and the disaster-tolerant design is established.Finally,simulation results verify the effectiveness of target recognition method and interpretation system.