Infrared Dim-Point Target Detection And Recognition Algorithm Under Complex Backgrounds

With the rapid development of stealth technique and successive service of stealth fighters, the IRAAM (Infrared Air-to-Air Missile) is facing a huge challenge from the Stealth Technique and Stealth Fighters to its development and combat operation. How to improve the detectable distance and detection probability of the IRAAM in stealthy combat environment in future has been a hot focus of researchers worldwide. Aimed to the difficulties of intense noise jamming、serious atmospheric attenuation and small imaging areas in the fields of infrared dim and small target, the dissertation mainly studies the detecting and recognizing technique which is applied to the multi-element IR Detector.Firstly, according to the feathers of infrared dim and small target, we study the characteristics of infrared target、sky background、noise. The multi spectral characteristics of typical targets are analyzed particularly, which build the foundation of follow-up research.Secondly, based on multi spectral characteristics of typical targets, Data cube and filter bank of background suppression are built. The improved algorithm of MHT combing with multi spectral characteristics is studied especially. The performance of detection and recognition with improved MHT algorithm is better than the classical one, including the detection and recognition probability, Computational complexityThirdly, by introducing multi spectral characteristics, the Single image is expanded to image cube. So the abundant Feature information, such as multispectral gray vector, motion information and spectral angle, which is beneficial for target recognitionFinally, we develop digital simulation and evaluation platform, Hardware-in-Loop simulation platform, which Provides effective Verification method for the development and evaluation of a new algorithm.