Weak Target Detection Based On Infrared Area Array Scanning System

IRSS (Infrared search system) has good invisibility, small blind area, fine anti-jamming ability, lightweight and small volume which becomes an important component of the modern defense systems. In order to increase the distance of detection, how to robustly detect the infrared dim-small targets under complex backgrounds have become more practical and challenging.In this paper, the main work is based on infrared focal plane array scanning system. Firstly, the infrared radiation of the typical invading targets and atmospheric transmission characteristics are analyzed to provide a theoretical basis for system. And then the range equation of infrared focal plane array scanning system is deduced to demonstrate the feasibility of it by comparing fully. In addition, the imaging features of targets and backgrounds in this system are introduced which lay the foundation for the research of detection algorithm.As a core part in full, this paper studied the related technologies within the framework of DBT (detection before track) to detect dim-small targets effectively. Firstly, a new background suppression algorithms conforming to infrared focal plane array scanning system are raised on the basis of the deep research on traditional background suppression which is called the SVD (singular value decomposition) method based on the limited adaptive neighborhood that is verified their respective advantages by simulation. Then, the target detection algorithm based on multiple algorithm fusion is proposed which can reduce background clutter and noise significantly by experiment. Next, in order to solve the uncontrolled problem of false alarm rate within the framework of DBT, the constant false alarm rate algorithm based on weighted feature of linear target is proposed which is verified the effectiveness by comparing. Lastly, some false targets are filtered by simple data association.In the end, the signal processing system based on FPGA and DSP is designed. The hardware design of relevant module on basis of the above detection algorithm above is in smooth operation. A series of experiments show the practicality of the algorithm proposed in this paper.