CNN 模板的设计及其在边缘检测上的应用

Design and Application in Edge Detection of Cellular Neural Network Template

本文提出了一种基于松弛—模拟退火算法的细胞神经网络的学习算法。该算法结合了松弛算法和模拟退火算法各自的优点在细胞神经网络模板设计的解空间内快速寻找最优解。将结果应用于图像的边缘提取上,计算机仿真结果显示,这种算法设计的细胞神经网络对图像边缘的提取取得了较好的效果。

Relaxation Anneal Aglorithm(RAA) for cellular neural network(CNN) is proposed in this paper.The cloning template can be obtained by using this algorithm.The algorithm help CNN get the best right solution for its template in its feasibal solutions.Simulation examples show that we can obtain better result than traditonary method in edge detec- tion of image processing with CNN method which cloning template is obtained by RAA.