摘要
通过结合自联想模型与异联想模型，提出了一种混合类间联想神经网络的光学模式识别系统。在该神经网络中，输入模式矢量不仅通过自联想识别自身，还通过异联想识别配对矢量，因而提高了识别概率。与匹配滤波器光学模式识别系统相比，识别结果直接。
By combining autoassociation and heteroassociation, an optical pattern recognition (OPR) system based on hybrid interpattern associative neural network is proposed. In this neural network, the input vector not only addresses itself by autoassociation but also addresses its mated vector by heteroassociation, so the correct addressing probability is increased. Compared with the matched spatial filter OPR system, the addressing results are direct, and this scheme is feasible and promising.
出处
《光学学报》
EI
CAS
CSCD
北大核心
1998年第10期1311-1316,共6页
Acta Optica Sinica
基金
中国科学院重点资助
关键词
混合类间联想
神经网络
光学模式识别
hybrid interpattern association, neural network, optical pattern recognition.
