摘要
针对数字图像放大问题,重点考虑了图像的两个重要特征:梯度和光滑性,提出了一种新的非线性偏微分PM-MF方法,并将其应用于图像放大。进一步从灰度偏移、信噪比、均方误差和梯度模值4个方面,提出了一套衡量图像放大质量的评价体系。实验结果表明该方法无论从保护边缘信息、灰度偏移以及图像清晰度等视觉特性方面均优于线性偏微分方法。
In view of image enlarge,a new method of PM-MF based on nonlinear Partial Differential Equation（PDE） is proposed and applied to image enlarge by considering the two most distinguishable characters：Gradient and smoothing degree.A new set of appraisal system for the qualities of image enlarge is proposed with farther research based on gray scale,ratio of signal to noise,mean square error and absolute value of gradient.From the result it shows that the proposed method precedes linear PDE whatever in keeping edge-based information,excursion of gray scale and the definition of pictures.
出处
《计算机工程与应用》
CSCD
北大核心
2010年第14期182-183,208,共3页
Computer Engineering and Applications
基金
陕西省自然科学基金No.2006F42
陕西省教育厅专项科学研究基金No.09JK796~~
关键词
图像放大
非线性扩散
评价体系
灰度偏移
image enlarge
nonlinear diffusion
appraisal system
excursion of gray scale
