基于整体变分模型的随机值脉冲噪声去除方法  

Removal Method of Random-Valued Impulse Noise Based on Total Variation Model

作　　者：胡樾 陶胜[2] HU Yue;TAO Sheng(Chengyi College,Jimei University,Xiamen 361021,China;School of Science,Jimei University,Xiamen 361021,China)

机构地区：[1]集美大学诚毅学院,福建厦门361021 [2]集美大学理学院,福建厦门361021

出　　处：《集美大学学报（自然科学版）》2023年第5期454-460,共7页Journal of Jimei University：Natural Science

摘　　要：为了有效地检测随机值脉冲噪声并去除干扰,提出一种新的开关整体变分去噪方法。根据不同密度的随机值脉冲噪声,分别在3 px×3 px、5 px×5 px和7 px×7 px的窗口邻域内进行噪声检测及检测结果修正,对判断为噪声的像素点采用基于整体变分模型的方法去噪,再采用动态阈值对去噪后的图像进行多次噪声检测和噪声去除,从而进一步提高去噪效果。仿真实验结果表明:对于不同密度的随机值脉冲噪声图像,该方法在有效去除噪声的同时还可以较好地保护图像的细节信息,峰值信噪比相比其他方法提高显著。To effectively detect random-valued impulse noise and remove its interference on the image,this paper proposes a new switching total variation denoising method.The method detects and modifies noise in different window neighborhoods of 3 px×3 px,5 px×5 px and 7 px×7 px according to different densities of random-valued impulse noise,and adopts a total variation based method to denoise the pixels that are finally judged as noise.In addition,the method uses a dynamic threshold to perform multiple noise detection and removal on the denoised image,thereby further improving the denoising effect.Simulation results show that the method can effectively remove noise and protect image details for images with random-valued impulse noise of different densities.It outperforms other methods in terms of peak signal noise ratio.

关 键 词：随机值脉冲噪声 噪声检测 整体变分模型 图像去噪 峰值信噪比 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]