基于OpenCV的图像滤波方法比较  

Comparison of OpenCV-based Image Filtering Methods

作　　者：赵博文 张力夫 潘在峰 王蓉[3] 郭雅馨 Zhao Bowen;Zhang Lifu;Pan Zaifeng;Wang Rong;Guo Yaxin(School of Communication and Information Engineering,Chongqing University of Posts and Telecommunications,Chongqing 400065,China;Hainan Branch of China United Network Communication Co.,Ltd.,Haikou Hainan 570216,China;Auto motive Electronics and Embedded Systems Engineering Center,Chongqing University of Posts and Telecommunications,Chongqing 400065,China)

机构地区：[1]重庆邮电大学通信与信息工程学院,重庆400065 [2]中国联合网络通信有限公司海南省分公司,海南海口570216 [3]重庆邮电大学汽车电子与嵌入式系统工程中心,重庆400065

出　　处：《信息与电脑》2020年第15期78-80,共3页Information & Computer

基　　金：大学生科研训练计划“V2X车联网系统及关键技术开发”。

摘　　要：视觉是人类最高级的感知器官,人来获取的信息83%来自视觉,图像为人类提供了很多至关重要的信息。数字图像处理是信号处理中的热门课题,在图像的获取和传输过程中,极易受到各类不同噪声的干扰,不同类型的噪声,其统计学特征存在差别,去除图像噪声最好的办法就是进行图像滤波,而处理不同的噪声就需要采用针对性的滤波方法。本文从椒盐噪声和高斯噪声入手,通过分析两种噪声的特点与原理,运用不同的滤波方法进行图像滤波处理的实验仿真分析对比,再通过MATLAB对去噪前后的图像进行相关灰度直方图的综合分析,最终得出去除椒盐噪声和高斯噪声最适合的方法分别是中值滤波和高斯滤波,并且随着滤波器阶数增大,噪声去除效果越好,但是带来的负面影响也会更大。Vision is the most advanced perception organ of human beings.83% of the information that people get comes from vision.Images provide a lot of vital information for human beings.Digital image processing is a hot topic in signal processing.In the process of image acquisition and transmission,it is easy to be interfered by various kinds of noise.Different types of noise have different statistical characteristics.The best way to remove image noise is to carry out image filtering,and to deal with different noises,we need to adopt targeted filtering methods.This paper starts with salt and pepper noise and Gaussian noise,through analyzing the characteristics and principles of the two kinds of noise,using different filtering methods to carry out the experimental simulation analysis and comparison of image filtering processing,and then through the comprehensive analysis of gray histogram of the image before and after denoising by MATLAB,it is concluded that the most suitable methods for removing salt and pepper noise and Gaussian noise are median filtering and Gaussian noise respectively Filtering,and with the increase of filter order,the better the noise removal effect,but the negative impact will also be greater.

关 键 词：高斯噪声 椒盐噪声 高斯滤波 中值滤波 OPENCV 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]