Space Target Infrared Imaging Simulation And Threat Assessment Technology

With the advancement of science,aircraft interception technology has developed rapidly,and the difficulty of aircraft penetration has greatly increased.Therefore,the aircraft needs to distinguish between the target and the decoy in time when the target strikes,and evaluate the threat of the target,which is conducive to improving the aircraft’s penetration capability.This paper studies the target threat assessment algorithm.The main tasks are as follows:Aiming at the problem of lack of corresponding data in infrared imaging of space environment targets,a simulation algorithm for infrared imaging of space environment targets is implemented to simulate the processes of target tracking and decoy separation.First,a variety of guidance methods are used to simulate the target tracking process,then the relative motion method is used to solve the trajectory of the decoy,then the radiation model is used to calculate the temperature of the target and the decoy,and finally the image is imaged according to the imaging principle.This paper uses simulation algorithms to generate a large number of simulated images,which provides effective data for the subsequent design of high-precision feature extraction algorithms for infrared image targets.Aiming at the problem that high-precision extraction of target characteristics is difficult due to the impact of point expansion and noise on the image plane when the target is flying,this paper proposes a high-precision target characteristics extraction method based on the square weighted centroid method and the discrete Chebyshev polynomial fitting method.First,the image is correlated,and then the centroid coordinates are obtained by the square weighted centroid method,and the edge coordinates are obtained by the discrete Chebyshev polynomial fitting method,and then the target size is obtained.The experimental results show that the centroid and size error extracted by the extraction method used in this paper is small,which meets the requirements of subsequent work.Aiming at the problem that actual physical values such as target speed and distance cannot be obtained under single-camera observation conditions,this paper proposes a method of fuzzy measurement of threat indicators according to the changes of target characteristics on the image,and proposes a method of using LSTM network to classify targets and decoys.method.The fuzzy measurement method establishes the relationship between the target characteristics and the number of frames,and the target characteristic change rate reflects the fuzzy measurement value of the actual physical quantity through the mapping function;the target and decoy classification method extracts the characteristics of the target and the decoy on the image sequence and enters the LSTM network Learn the features,and finally output the result according to the classification probability.The experimental results show that the classification algorithm has a better classification effect on samples.Aiming at the low accuracy and slow speed of traditional target threat estimation methods,this paper proposes a target threat assessment algorithm based on improved PSOLSSVM.First,each group of threat indicators is scored by experts to construct a data set,and then the improved PSO algorithm is used to determine the initial parameters of LSSVM,and finally LSSVM is used to train and verify the data set.Experimental results show that the root mean square error of the test set is as low as 0.1160 when the algorithm is faster.