基于扩散张量的图像混合去噪方法  

Image Mixed Denoising Method Based on Diffusion Tensor

作　　者：陈龙[1] 

机构地区：[1]昆明理工大学理学院,昆明650500

出　　处：《价值工程》2015年第9期325-327,共3页Value Engineering

基　　金：云南省自然科学基金(2011FZ025);昆明理工大学人才基金(2008-72)资助

摘　　要：图像去噪过程中,主要目的是有效去除噪声的同时能较好地保存图像的边缘与细节。基于此,本文根据自蛇模型与相干增强扩散的特征,提出了相应的混合去噪方法,并加入了保真项。实验结果表明该方法能有效去除图像噪声,同时对边缘信息的保留也较好。In the process of image denoising, main purpose is to remove noise effectively and preserve well edges and detail information. Based on this, according to characteristic of conherence enhancing diffusion and self-snake model, an novel image denoising method based on these two model is proposed, and a fidelity term is added. Experiment results show that the proposed method not only remove noise efficiently, but also keep better for edge information.

关 键 词：图像去噪 扩散张量 自蛇模型 各向异性扩散 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]