本文提出一种模板可编程细胞神经网络的硬件实现方法 ,设计构成CNN的细胞体电路、A模板电路和B模板电路 ,组成CNN并进行在图像处理中的应用研究 .仿真结果表明 ,所设计的硬件电路具有结构简单、功耗低、频率特性好、模板参数可编程等特点 ,可以方便地构成各种规模的CNN ,在图像处理应用中具有一定的灵活性和通用性 . In this paper, we propose a hardware implementation method of template programmable cellular neural network, design CNE cell body circuit, A template circuit and B template circuit to form CNN and apply it in image processing.The simulation results show that the designed The hardware circuit has the advantages of simple structure, low power consumption, good frequency characteristics and programmable parameters of the template, which can conveniently form CNNs of various scales and has certain flexibility and versatility in image processing applications.