基于形态学和坐标变换的二维CA图像处理技术研究

A New CA Method for Image Processing Based on Morphology and Coordinate Logic

深入研究了一类二维细胞自动机(CA),提出了一种用于二值和多值图像处理的二维CA实现方法。对于二值图像,提出了类似于数学形态学腐蚀、膨胀的CA布尔逻辑规则;对于多值图像,采用坐标逻辑运算把图像分解成多个二值图像后用CA并行处理。所有的CA规则都可以用细胞神经网络(CNN)实现。最后,将这种方法用于图像滤波,获得了很好的实验结果。

In this paper, a new kind of two-dimensional cellular automata is investigated detailedly, and a new CA-based method for image processing is proposed.For binary image, the CA-based erosion and dilation rules like morphology are found. However,multi-value image is decomposed to several binary images with coordinate logic, and then be processing with CA. All the CA rules can be implemented with Cellular Neural Network (CNN). At last, this approach is used to image filtering and good experimental results are obtained.