您的位置:首页 > 论文页面
图像去噪的保细节各项异性扩散模型
发表时间:2015-01-15 浏览量:2107 下载量:660
全部作者: | 杨莹莹,何传江,王伟 |
作者单位: | 重庆大学数学与统计学院 |
摘 要: | 为有效去除图像噪声,同时保留图像的纹理、细节和弱边界等信息,基于改进的Perona-Malik模型和水平集演化理论,提出一个新的各项异性扩散模型。该模型的偏微分方程由一个扩散项和图像自适应保真项组成。扩散项利用归一化方差的加入而使模型较好地去除图像噪声,图像自适应项利用扩散系数的改进而使去除噪声的同时锐化图像的边缘信息,保留更多的细节。实验结果表明:该模型不仅能够很好地去除噪声,同时能够更好地保留图像的重要细节。 |
关 键 词: | 计算数学;图像去噪;各项异性扩散;曲线演化;水平集 |
Title: | Anisotropic diffusion model for image denoising with detail-preserving |
Author: | YANG Yingying, HE Chuanjiang, WANG Wei |
Organization: | College of Mathematics and Statistics, Chongqing University |
Abstract: | To effectively remove the noise while keeping image features such as details, textures and weak edges, this paper proposes a new anisotropic diffusion model for image denoising based on the improved Perona-Malik model and curve evolution via level sets. The partial differential equation for the proposed model consists of an anisotropic diffusion term and adaptive image fidelity term. The anisotropic diffusion term uses the normalized variance that can remove the noise effectively, and the adaptive image fidelity term includes improved diffusion coefficient which can sharpen the edge of image and keep more details while removing the noise. Experimental results show that the proposed model not only can remove the noise effectively but also can preserve important details of images. |
Key words: | computational mathematics; image denoising; anisotropic diffusion; curve evolution; level set |
发表期数: | 2015年1月第1期 |
引用格式: | 杨莹莹,何传江,王伟. 图像去噪的保细节各项异性扩散模型[J]. 中国科技论文在线精品论文,2015,8(1):12-18. |
1
评论数 0
请您登录
暂无评论