Research On Detection And Tracking Algorithm Of Infrared Small Target Based On Human Visual System

Infrared target detection technology is widely used in military,national defense and other fields.With the development of computer technology,the detection and tracking accuracy of infrared target is getting higher and higher.Infrared target detection and tracking technology is applied in more fields such as industry,agriculture and transportation.Among them,the detection and tracking technology of infrared small and weak targets is difficult and of high military and civilian value,which has become a popular study widely concerned by researchers.On this basis,the research of infrared dim small target detection and tracking algorithm based on human visual system is developed in this paper.Firstly,as for infrared image,we analyze the characteristics of image background,noise and dim small targets,and introduce infrared image preprocessing method based on background suppression and noise elimination,which provides the theoretical basis for the research content of this paper.Secondly,aiming at detecting false alarms caused by background edges and high-brightness corners,an improved saliency detection method is proposed based on the human vision characteristics and the orientation characteristics of small targets and background edges in the local area.Aiming at the problem of unstable detection effect when detecting targets with different scales,combining significance detection and scale space,a scale adaptive dim small targets detection method is proposed.The effectiveness of the method is verified by experiments.Then,we study the method about infrared small target detection and tracking based on eye movement mechanism.This method first detects the significance of infrared image,then determines the region of interest,detects the target in the region of interest,and finally tracks the target by simulating the eye movement mode of human visual system.Aiming at the problem that the tracking effect of the PID tracking algorithm based on the eye movement mechanism is unstable due to the fixed parameters,a small target tracking algorithm that can be automatically corrected according to the position error is proposed,and a comparison experiment between the measured and simulated scenes is conducted to verify the accuracy of the method.Finally,based on the random set theory,the pre-detection tracking method of multiinfrared dim small targets is studied.Based on the characteristics of infrared image and multi-target motion,and on the basis of Me MBer filter,a multi-infrared dim small target detection and tracking method based on significance detection and new target automatic search is proposed.First,infrared image sequences are detected for significance to obtain the significance image sequences.Then,we use SCK-Me MBer-TBD algorithm and new target automatic search algorithm to detect and track infrared small targets.Through comparative experiments and analysis,the real-time performance and stability of the algorithm are verified.