Research On Feature Analysis And Recognition Strategies Of Dim Small Targets Under Complex Background

Due to the improvement of the military strength and the complexity of the combat environment in the world,the infrared detection system is widely used in the military and civil fields such as precision guidance,aerospace and so on.In order to improve the detection distance as much as possible and detect the moving target in advance,the radiation energy of the target in the original image is often very weak.And the target is in the form of dot,lacking of texture information.At the same time,the interference of complex ground background makes the target easy to be lost.Based on these objective factors,infrared small target detection under complex ground background has been a hot and difficult issue in current research.Based on the analysis of the characteristics of dim small targets,background and noise in complex background,a stable and invariant feature knowledge base of dim small targets in complex background is established in this paper.On this basis,this paper formulated front-end image acquisition strategy and back-end image processing algorithm,which effectively suppressed the false alarm interference of high-noise ground background and realized the stable detection,identification and tracking of small targets in complex ground background.The main work and contributions of this paper are as follows:1.Based on the special application background of complex ground,this paper researched the dim small target detection methods proposed in domestic and foreign literature.According to the special application background and characteristics of complex ground,the corresponding framework of dim small target detection was proposed.2.The characteristics of infrared small targets were summarized.In addition,the characteristics of background and noise in original images were also deeply explored,and an effective distinguishing method is proposed.By summarizing the characteristics,this paper established a complete and stable knowledge base of target characteristics,and solved the problem of false alarms and unstable track caused by complex ground background.3.Based on the study of human vision system,a fused filtering method and an adaptive segmentation algorithm based on complexity were proposed.These methods solved the problem that the target is too weak and the detection rate is too low in the complex background.4.Based on the study of the time domain characteristics of sequential images,a new image acquisition strategy is used to expand the field of view and highlight the motion of the target.At the same time,according to the feature continuity of the target in the time domain,the false alarm removal mechanism and the pipeline filtering algorithm based on the target feature knowledge base are proposed,which removed the false alarms caused by the complex background.This paper tested a large number of data sets collected from real scenes and simulated.The experimental results showed that the strategy is correct.The problem of false alarm and missed detection caused by strong noise interference in complex ground background is effectively solved.