用HDL语言描述和模拟图像二值化神经网络

Describing and Simulating A Neural Network Model for Image Binarization Using HDL

神经网络专用电路实现是目前神经网络实现研究的主要方向。本文基于我们所提出的数字式细胞神经网络．针对它在数字图像二值化中的应用，采用硬件描述语言对这个专用电路进行描述和模拟。在系统设计中，采用了微程序控制方法和流水线技术。仿真结果表明了硬件实现算法的正确性和可行性，同时也为神经网络的实现打下了良好的基础。

Appication specific integrated circuits implementation of neural metworks is an important aspect of research on artificial neural network.This paper gives a description of such a circuit to implement image binarization using our degital celiular neural network.In the system design,we adopt microcode control approach and pipeline processing,so as to make the time and space complexity low.Simulation results show true results and feasibility of using hardware to implement the algorithm and at the algorithm and at the same time,it gives a design basis for VLSI implementation of meural metworks.