Millimeter-wave Infrared Composite Sensor Signal Processing System

For target recognition, millimeter wave sensor and infrared sensor have their own unique advantage. Millimeter wave sensor can distinguish between metallic target and surroundings effectively, and infrared sensor can easily identify thermal target. Combination of them will increase the probability of target identification, and improve positioning accuracy. The signal processing system of the composite sensor is discussed in this paper, including target recognition and target positioning.In this paper, the principle of target recognition using millimeter wave sensor and using infrared sensor are first described and the approximate mathematical model of their output signal is presented. According to the model, the characteristics of output signal are analyzed. Then the problem of sampling, preprocessing and extracting feature vectors are discussed. The using of Infrared sensor arrays is mentioned to solve the disadvantage of the identification of target size. The recognition methods discussed in this paper include minimum distance classification and fuzzy classifier method, and the performance of these two methods are studied and analyzed. Then the D-S evidence theory data fusion algorithm is discussed and analyzed. At last of this paper, using data fusion to improve target positioning accuracy is expounded.