Color Recognition System by Optical Neural Network

本文报道了两种用于颜色识别的光学神经网络。第一种是基于计算机辅助识别系统建立的颜色系统，通过仿真和硬件系统验证颜色识别。另一种是基于灵长类眼脑模型。仿真结果证实输出与大脑皮层第四视觉区（V4）匹配。发现其隐层单位与猴神经节细胞相似。通过对该系统的改进，有望为眼脑的解释和类似人类的颜色识别提供支持。

This paper reports on two types of optical neural network for color recognition. The first type is based on the color system established by the C.I.E.Color recognition was confirmed by simulation and hardware system of this type.The other type is based on the Primates ophthalmencephalon model. As a result of the simulation, the output matching the fourth visual areas (V4) of cerebral cortex was confirmed. The hidden layr units were found to be similar to the ganglion cell of monkey. By improving this system, supports on explication of the ophthalmencephalon and the color recognition similar to human are expected.