Research On Detection Algorithms Of IR Small Weak Target

Along with the development of modern science and technology, and the requirement of modern war, as the big advantage of IR search and track system (IRST), the IR search and track system become the most important precise guide technology.Recently, the IR small weak target detection algorithm could be mainly divided in to two classes, Detection Before Track Approach and Track Before Detection Approach. The former one is much simple and easy to realize, but high signal to noise ratio (SNR) is required. This method is mainly discussed and researched in this paper.As far as the Detection Before Track method is concerned, single frame processing which aim to get the set of the candidate targets is the most important procedure and it would influence the performance of the whole method in a large degree. Up to now, most of the single frame detection methods are good enough to handle the normal SNR images in which the backgrounds are smoothly distributed. But when used to handle the heavy clutter images, their performances are no longer guaranteed.First, this paper introduces a new algorithm based on single frame detection, and its performance is proven to be stable and the complexity is low.Then, based researching on basic Multiple Hypothesis Test related algorithms, considering the background of IR small weak target detection, an improved and simplified MHT algorithm is introduced and it can reduce the false alarm rate by a large degree. Finally, software platform for IR detection and track is briefly introduced, and then it is used to evaluate the algorithms mentioned in this paper.