THREE-LAYER OPTICAL NEURAL NETWORK FOR PATTERN RECOGNITION WITH ROTATION INVARIANCE
具有转动不变性模式识别能力的三层光学神经网络模型

A model of three-layer optical neural network is proposed for pattern recognition with rotation invariance. As an example, the model is applied to recognize the projective views of four classes of aircrafts. Investigation of computer simulations shows that the model has good capacity of pattern recognition for patterns both inside and outside the training sets, as well as partially shaded patterns.