运用四元数神经网络模型识别16色彩色字符

USING THE HAMILTON NEURAL NETWORK MODEL TO RECOGNIZE THE 16-LEVEL COLOR CHARACTERS

本文把四元数离散神经网络模型应用于十六色彩色中英文字符的识别中,讨论了彩色文字识别中的一些特点,如由于颜色的改变,导致信噪比发生变化,所以网络的存贮容量和纠错能力也随之而改变.

We apply the Hamilton discrete neural network to recognize 16-level color characters, and discuss some features of the model. If we change the color of the characters, the signal-noise-ratio is changed, so the storage capacity and the error tolereance capacity are also changed.