文献类型：期刊文献

中文题名：一种新型的图像噪声滤波算法研究

英文题名：Research on Image Noise Filtering Algorithm Based on Improved Pulse Coupled Neural Network

作者：孙建召[1];常青[2]

第一作者：孙建召

机构：[1]河南经贸职业学院信息管理系;[2]新乡学院计算机与信息工程学院

第一机构：河南经贸职业学院信息管理系,郑州450018

年份：2016

期号：6

起止页码：848-851

中文期刊名：控制工程

外文期刊名：Control Engineering of China

收录：北大核心:【北大核心2014】；CSCD:【CSCD2015_2016】；

基金：河南省科技技术厅项目(941202015Y1877)

语种：中文

中文关键词：模糊神经网络;彩色图像;隶属函数;鲁棒性

外文关键词：Fuzzy neural network;color image;membership function;robustness

摘要：为了提高图像质量,有效的滤除被污染图像中的噪声,提出了一种具有鲁棒性的模糊神经网络(Robust Fuzzy-Neural Network,RFNN)图像噪声滤波算法。首先,对RFNN进行了网络优化训练,然后利用优化后的网络对含有噪声的彩色图像进行滤波处理。在RFNN的构造中,设计了一个新的隶属函数,从而使该算法在彩色图像噪声处理过程中具有鲁棒性。最后的仿真实验结果表明,相比传统的滤波算法,该新型滤波器具有更好去噪能力,并且在去除噪声的同时能够很好的保留图像的边缘和细节等信息。
In order to improve the quality of the image, the noise of the contaminated image is effectively filtered, and a robust fuzzy-neural network(RFNN) image noise filtering algorithm is proposed. At first, the network optimization training of RFNN is carried out, and then the optimized network is used to filter the color image containing noise. In the structure of RFNN, a new membership function is designed, which makes the algorithm robust to the noise processing of color images. Finally, the simulation experiment results show that compared with the traditional filtering algorithm, the new filter has better denoising ability, and can preserve very well for the edges and details of the image while removing the noise.