基于四阶微分全变差图像去噪模型.PDF

第26卷 第3期 计 算机 技 术 与发 展 Vol.26 No.3 2016年3月 COMPUTERTECHNOLOGY AND DEVELOPMENT Mar. 2016 基于四阶微分全变差的图像去噪模型 王 璐 (昆明理工大学理学院,云南 昆明 650500) 摘 要:针对现有全变差(TV)方法效果不太理想,在去除噪声的时候不能较好地保护图像的边缘信息,而且恢复图像易出 现“阶梯效应”和“块效应”的问题,文中提出了一种基于四阶微分全变差的图像去噪模型。 首先论述了传统全变差模型去 噪方法及其高阶微分方程去噪方法各自的优缺点;然后将带有边缘指标的全变差模型与四阶微分理论相结合,得到了一 个新的带有边缘指标的自适应全变差去噪模型,并引入差分方程去定义模型中的变量。 实验结果表明,该模型能较好地 抑制噪声、保留边缘特征和衰减图像的“阶梯效应”,并能较好地避免图像的“块效应”。 关键词:图像去噪;全变差;四阶微分;边缘保真;块状效应 中图分类号:TP391.41 文献标识码:A 文章编号:1673-629X(2016)03-0085-04 doi:10.3969/j.issn.1673-629X.2016.03.020 Image De-noising Model Based on Total Variation of Fourth-order Differential WANG Lu (Faculty of Science,Kunming University of Science and Technology,Kunming650500,China) Abstract:The effect of the existing Total Variation (TV) method for image denoising is not ideal while removing noise,it is not well protection oftheedgeinformationofimageandeasytoappearthe staircaseeffect and blockyeffect whenrecoveringtheimage.Ai- ming at thisproblem,a new method of image denoising based on TV of fourth-order differential isproposed in thispaper.First,the ad- vantages anddisadvantagesofthetraditionalimagedenoisingmethodsofTVandhigher order differentialequationsarediscussedrespec- tively.Then,combined thedenoisingmodelofTVwithfourth-orderdifferentialtheory,anewadaptiveATVmodelwiththeedgeindica- tor isobtained,and a rationaldifferentialmaskineightdirectionsisdrawn.Theexperimentalresultsshowthatthismethodcanreducethe noise well,and preserve edge featuresbetter,and reduce the influence of the staircase effect,avoiding the blocky effect of the im- age. Key