Research On Method Of Infrared Target Detection And Bearings-only Guidance

In the modern warfare,precision-guided weapons have to face more severe challenges due to the high-intensity electronic countermeasures and the complex and variable battlefield environment.Therefore,it is particularly important that precision-guided weapons have the ability to resist electronic jamming,accurately identify and track targets in complex environments,and have all-day,all-weather combat capabilities.Infrared imaging guidance,as a passive guidance technology,has the characteristics of strong adaptability,full-time work,and penetrability compared to visible imaging guidance.Compared with active guidance such as radar and laser,it has better concealment and anti-jamming ability.As the target moves from far to near,the image of the target in the imaging plane changes gradually from a point target to a plane target.Because of the interference of the complex background environment,the signal of the point target is usually overwhelmed by noise.Therefore,the long-range detection capability of the missile can be effectively improved if the point target could be precisely detected from the complex environment.The plane target is usually corrupted by noise,angle of view,resulting in a large difference between the target and the typical template.Hence,the research of target matching algorithms under the influence of various interferences can effectively enhance the target matching accuracy and improve the guidance precision.The valid information provided by the infrared imaging seeker is only the target azimuth information,as a consequence,how to estimate the target position accurately under bearings-only information becomes the key to improve guidance precision.Based on the analysis above,the main research contents of the paper are as follows:1)Based on the characteristics of frequency-domain and time-domain,the detection method of infrared small dim target in complex background is studied.Firstly,the difference of target,noise and edges in frequency-domain is analyzed in detail,and a target candidate region detection method is proposed based on the non-directionality and local high energy characteristics of the small dim target.Then,a unified time-domain profile model is established by taking the target as a special noise in the infrared image,and a predictor-corrector algorithm is proposed to predict the reference profile of the time-domain profile,so as to achieve the goal of accurate detection of the target position.2)In order to solve the problem of plane target detection based on template matching,firstly,the concept of main gradient direction is introduced into the log-polar transform,and the invariance of log-polar transform to rotation,translation and scaling is realized.The stability of main gradient direction is verified by simulations.Secondly,from the statistical point of view,the characteristics of image region are analyzed.Based on the variance of image region,an adaptive fast matching strategy is proposed.Compared with the ergodic search,simulation results show that the matching strategy is fast and effective.3)Based on the shape matching algorithm,the problem of plane target matching under the influence of occlusion and deformation is studied in detail.Firstly,the problem of over evolution in the method of discrete curve evolution is analyzed.Based on the variance of the correlation sequence,an improved method of discrete curve evolution is proposed.Secondly,a set of multi-scale invariant features is designed to enhance the stability of target feature description,and the problem of target matching is transformed into the problem of feature classification.Finally,by analyzing the result of feature matching,taking the previous matching result as the prior information of the next matching,a circular Smith-Waterman algorithm is proposed,and the effectiveness of the algorithm for local shape matching under the interference of occlusion and deformation is verified by simulations.4)The infrared imaging guidance under enhanced target observability is studied.Firstly,the relative motion models of missile and target based on rectangular coordinate system and modified polar coordinate system are established.Then,the observability of the target state under bearings-only information is analyzed in detail,the observability condition is given,and its physical meaning is explained.Finally,taking the size of the uncertainty area of the target state estimation error as the measurement of observability,the performance index considering both observability and guidance accuracy is designed,and the analytical form of the missile guidance law is deduced.The effectiveness of the guidance law for the target state estimation is verified by simulations.5)Based on the digital simulation system of infrared imaging guidance,the influence of the target observable term on the guidance accuracy is analyzed and verified.Firstly,a digital seeker model considering various error factors is established to simulate the target detection results,and corresponding guidance strategies and data validity criteria are designed for different working conditions of the seeker.Secondly,the guidance law considering the target observability enhancement is combined with the leading angle + proportional integral guidance law.Aiming at the influence of the observable term in the guidance law on the guidance accuracy,the standard six-degree-of-freedom trajectory simulation and Monte Carlo simulation are analyzed in detail.