基于引导图像的边缘噪声滤波算法  

Image Edge Noising Filtering Algorithm Based on Guided Image

作　　者：高红霞[1] 

机构地区：[1]河南工程学院计算机学院,郑州450007

出　　处：《科学技术与工程》2014年第9期78-81,共4页Science Technology and Engineering

基　　金：国家自然科学基金(71173248)资助

摘　　要：由于通常的邻域运算会改变图像边缘点的灰度值,使图像的边缘变得模糊,为了改善这一现象,提出了一种基于引导图像的边缘噪声滤波算法。该算法由局部线性模型推导而来,将原始图像或其他变换形式定义为引导图像。通过对引导图像进行分析,并调节正则化参数,利用引导图像掩模对图像的边缘进行平滑处理,有效地去除了噪声。通过与其他四种常用的滤波算法进行对比实验,表明该算法的均方误差MSE仅为0.0015,峰值信噪比PSNR为28.26,远远优于其他四种常见滤波算法,不仅对图像进行了平滑去噪,在很大程度上还保护了图像的边缘信息。Due to the gray value of image edge is changed by the usual neighborhood computing, which makes the image edge to be fuzzy. To improve this phenomenon, the image edge noising filtering algorithm based on guided image is proposed. The algorithm is derived from the local linear model, and the original images or other transformation forms are defined as the guide images. The guided image mask is used to smooth the edge of image by analyzing the guided image and adjusting the regularization parameters, which effectively removes noise. The comparative experiments with other four common filtering algorithms show that the mean square error of the proposed algorithm is only 0. 001 5 and the peak signal noise ratio is 28.26, which are far better than the other four common filtering algorithm, not only smoothing and de-noising for the image, but also protecting the edge information of the image largely.

关 键 词：图像滤波 引导图像 边缘去噪 正则化参数 高斯噪声 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]