%0 Journal Article
%T 基于高斯-中值的钢轨表面缺陷图像滤波研究<br>Image filtering of rail surface defects based on gauss-median
%A 顾桂梅
%A 
%A 
%A 冉建民
%A 周咏&lt
%A br&gt
%A GU Guimei
%A 
%A 
%A RAN Jianmin
%A ZHOU Yong
%J 铁道科学与工程学报
%D 2018
%X 针对传统滤波算法在钢轨表面缺陷检测中噪声滤除效果的缺点，提出一种高斯-中值滤波算法。将图像反转，使缺陷及一些被氧化处与正常的钢轨表面的灰度亮度发生反转。对反转后图像的滤波模板窗口求加权平均值，将图像中的每个点的像素灰度值与其加权平均值进行比较，若该点的像素灰度值大于其加权平均值，则用中值滤波算法进行处理，否则用高斯滤波算法进行处理。将仿真结果与传统的方法相比较表明：该方法去除噪声效果更好，并能很好的保护图像细节和改善缺陷处的边缘细节。<br>A Gaussian-median filter algorithm was proposed for the shortcomings of traditional filtering algorithm for noise filtering in surface defect detection. The image was reversed, so that some of the defects and oxidation can reverse normal rail surface of the gray-scale brightness. After the inverse image of the filter template window weighted averaging, the pixel gray value of each point in the image was compared with its weighted average pixel gray value. If the pixel gray value of the point is larger than its weighted average value, the median filter algorithm was utilized for processing, otherwise the Gaussian filter algorithm was used for processing. Compared with the traditional method, the simulation results show that the method has better noise removal effect, and it can protect the image details as well as improve the edge details
%K 钢轨表面缺陷
%K 图像
%K 噪声
%K 滤波
%K 图像反转&lt
%K br&gt
%K rail surface defect
%K picture
%K noise
%K filter
%K image reversion
%U http://www.jrse.cn/paper/paperView.aspx?id=paper_318317