|
|
基于变换域的图像去噪声
摘要
在日常生活中，图像已经成为人们获取、传递信息的重要手段，但图像在产生、传播时容易受到噪声的干扰，使得图像变得模糊，因此对模糊图像进行去噪处理变得很重要。
本文主要研究中值滤波和小波变换来去除图像中脉冲和高斯噪声，研究内容主要分为以下几个部分：
一、本文研究了利用标准的中值滤波、极值中值滤波和极值中值滤波的改进算法来去除图像中的脉冲噪声，通过实验比较，改进算法具有较好的去噪效果。
二、本文研究了利用基于阈值的小波变换来去除图像中高斯噪声，研究了软硬阈值处理函数和新阈值处理函数，通过实验比较，采用新阈值处理函数的小波去噪效果较好。
三、本文针对单一去噪算法在去除混合噪声的不足，研究了将两种改进算法结合来去除混合噪声，通过实验比较，结合算法去噪效果好。
关键词：图像去噪 中值滤波 小波去噪
The Research of Image Denoising Based on Transform Domain
Abstract
In daily life,picture has become an important way for people to obtain and transmit information.However,pictures,especially during the process of it’s generation and transmission, can be easily interrupted by noise,as a result,it becomes vague.Therefore,it is important to denoise the vague pictures. The text will mainly study the using of the median filtering and wavelet transformation to remove pulse and Gaussian noise in a picture.The content of the study will be given below.
I.The text studied the removal of pulse noise in a picture,by adopting the standard median filtering,extremum median filtering algorithm and the improved one that based on extremum.According to the experiment,the text also finds out that the improved algorithm has a good effect on denoising.
II.The text studied the removal of Gaussian noise,soft-hard processing threshold and new threshold processing function,by adopting wavelet denoising that based on threshold .According to the experiment,wavelet denoising that based on new threshold processing function has a good effect on denoise.
III.Aiming at the deficiency in removal mixed noise of single denoising algorithm,the text explored the method that combined the two improved algorithm to remove mixed noise.According to the experiment,combined algorithm has a good effect on denoising.
Keywords ：Image denoising;Median filter;Wavelet transform
目录
