文章基本信息

标题：Effect of Regularization parameter on Total Variation based denoising of Magnetic Resonance Images
本地全文：下载
作者：Nikita Joshi ; Sarika Jain ; Amit Agarwal 等
期刊名称：International Journal of Advances in Soft Computing and Its Applications
印刷版ISSN：2074-8523
出版年度：2018
卷号：10
期号：3
出版社：International Center for Scientific Research and Studies
摘要：Noise affects Magnetic Resonance (MR) images, due to which the problem of inaccurate medical diagnosis occurs. Therefore noise removal is an important task while dealing MR images. In this paper, the discrete total variation method has been discussed and analysed for removing noise from Magnetic Resonance Images. The effect of regularization parameter lambda has been studied for denoising. This method has been extensively experimented with MR images by varying the parameter lambda. The evaluation metrics are Peak Signal to Noise Ratio (PSNR) and Mean Square Error (MSE). The experiment demonstrated that the value of PSNR decreases and MSE increases as the value of lambda increases from 0.01 to 1.0. The noise is reduced and contrast is improved.
关键词：Discrete total variation; image denoising; magnetic resonance images; regularization parameter