人工神经网络技术在目标识别中的应用

在多传感器信息融合的基础上讨论了目标种类识别的有关问题，并给出了直升机、固定翼飞机、导弹、水面舰艇、鱼雷、潜艇和水雷的识别准则和方法，并在此基础上提出了基于BP神经网络的目标种类识别模型。仿真计算结果表明，只要学习样本足够多，能够有效保证识别的正确性和克服基于规则的识别方法在模糊区识别效果差的问题。

Application of Artificial-Neural-Network-Based on Target Recognition

This Paper discusses Target Type Identification issues based on the concept of MultiSensor Information Fusion. And presents identification methods & rules for Helicopters, fixed-wing aircrafts, missiles, surface ships, torpedoes, submarines as well as mines. Target Type identification model is proposed based on the Back Proprogation Neural network theory. The simulation result shows the proposed method can effectively ensure identification accuracy and avoid poor Identification results in the area of uncertainty by the identification method based on the rules theories as long as the study samples is enough.