Abstract:For this topic we try to find a general and practical method. Due to its strong capabilities of self-organizing, self-learning and fast parallel processing, neural network is expected to solve this problem. Afterconsidering the main disadvantages in back-propagation neural network: long-training and local-minimum problems, we propose an architecture of hierachically connected network. As a result, the trainingbecomes much faster and the local-minimum much scarcely appears. Some satisfactory results in the attitude measurement of aircraft have been obtained.
[1] Brooks R A. Symbolic reasoning among 3-D model and 2-D images. AI, 1981; 17: 285-340
[2] 王建刚,朱枫,姚筱亦.三维物体识别及姿态测定的约束推理系统.第3届青年机器人学术会议,桂林.,1990.9
[3] Dydani S A et al.Aircraft identification by moment inviavant. IEEE Trans Computer 1977; C-26: 39-46
[4] Klansky J S, Davision G A. Recognizing 3-D objects by their silhouettes. J of the Society of Photo-Optical Instrumentation
[5] 大津展之.N次自己相关マスクにと图形特征抽取.信学技报,1978:78-31
[6] 马颂德.神经网络分类器.神经网络及其应用学术讨论会,北京,1989