金字塔结构逻辑运用二值脉冲对简单图形处理

本文根据元胞自动机模型划分方法, 将二维图像分解为2×2矩阵单元结构. 提出了几种逻辑运算式, 用以分类由黑白二值点构成的2×2矩阵图形. 通过CNN神经网络的多层结构形式, 分析了金字塔结构逻辑在相似的组合形式下, 对二值图形边缘检测和池化的功能. 通过同步脉冲形式能将灰度图像, 分解为多个时间维度的二值图形, 方便多层金字塔逻辑运算处理. 分析了如何采用延时继电器使金字塔结构逻辑具有记忆的特性. 讨论了3×3输入金字塔模型, 在不规律脉冲情况下, 通过逻辑运算对线性交点检测的可能.

Pyramid Structural Logic Using Binary Pulse for Simple Graphics Processing

In this paper,two-dimensional image is decomposed into 2×2 matrix unit structure according to the method of cellular automata model partition.Several logic operations are proposed to classify 2×2 matrix graphs composed of black and white binary points.The functions of edge processing and pooling of binary graphics in the similar combination form of pyramid structure logic are analyzed through the multi-layer structure form of CNN neural network.The gray scale image can be decomposed into binary graphics with multiple time dimensions by synchronous pulse form,which is convenient for multi-layer pyramid logic operation.This paper analyses how to use delay relay to make pyramid structure logic have the characteristics of memory.This paper discusses the possibility of detecting linear intersection points by logic operation under irregular impulses based on the 3×3 input pyramid model.