Low Altitude Target Attribute Recognition Based On Multi Sensor Information Fusion Technology

As one of the key aspects of modern warfare, the ability to correctly identify the properties of low altitude targets have a significant impact on the distribution and deployment of fire and the entire battlefield situation. Therefore, effective target recognition technology has important significance for Air-Defense War. In this paper, based on radar-infrared sensor information fusion technology, to propose a RS-RBF low-altitude target recognition method,achieved the aim to enhance the recognition rate target..Firstly,this paper introduces the purposes, significance and research status of multi-sensor information fusion technology at low altitude target attribute identification Based on it, the paper determines the level of multi-sensor information fusion, selects the structure of information fusion, and introduces the track correlation of target recognition and some of the existing target recognition algorithm.Secondly,this paper designs low altitude target recognition models based on BP neural network and RBF neural network. The six attribute of low altitude target to be recognized as the input vector of the two kinds of identification model, type of the object as the output vector model be trained and tested.The simulation results show that the RBF neural network has more advantages than BP neural network.Then to study the rough set theory and design a model recognition based on RS-RBF. By using attribute reduction of decision table technique to reduce the properties of low altitude target, achieved the purpose of removing redundant information. The attribute after reduce as input and target type as output of the RBF network to train and test the network. Through the experimental comparison, RS-RBF method is more effectively than the RBF network model in target recognition.Finally, researching the Radar and infrared sensor information fusion system based on the least square method. Based on it, detection of angle and position information by infrared sensor and radar to information fusion. The fused data as input vector of RS-RBF object recognition model, to distinguish low altitude targets, such as the armed helicopter, bombers and early warning aircraft. The simulation results show that low altitude target attribute identification based on multi sensor information fusion rate is higher, the stability of the system is stronger than a single sensor target identification.