Research Of Composite Recognition Technology Based On Millimeter-wave Infrared Target Feature

In order to acquire a high hit and damage rate of conventional ammunition in the increasingly harsh modernized battle field, many countries develop the terminal sensitive projectile(TSP) rapidly. TSP sensor system typically employs millimeter-wave/infrared(MMW/IR) sensors to detect and identify targets, but sometimes the target information of composite detectors is imprecise, even contradictory. Therefore it is of vital importance to analyze a composite target recognition algorithm, using the characteristics of MMW/IR sensors data, to improve the target identification capability, hence enhancing anti-jamming capability to better determine the target center, turn on the ignition circuit and fire the EFP to complete the attack mission.This article mainly focuses on a MMW/IR signal processing and multi-sensor data fusion detection technology. According to the system requirements, the paper first elaborates the basic principles of MMW/IR channels, and then estblishes a passive infrared and passive millimeter wave complex detection system, thus fusing the dual-mode selection information with feature layers. Secondly, it carries modeling and simulation out of the characteristics of MMW/IR channels, and analyzes their waveforms. Thirdly, it captures the signal characteristics by using the wavelet transform technology which completes signal preprocession and decomposition, noise reduction and reconstruction. Finally, the system identifies a Neural Network model to complete the single target identification, which owns much to the neural network technology, and sets up a promoted data fusion model, with using the D-S Fuzzy Recognition, to produce a comprehensive verdict. Through these technologies, the system has greatly improved the target recognition ability of complex identification system, compared to the single probe system.