摘要
恢复含乘性噪声的图像是当前图像处理的重要研究课题.提出基于积分微分方程的乘性噪声去除模型.新模型中假定乘性噪声服从伽马分布.讨论了经典的去除乘性噪声的AA模型,在此基础上通过在逆尺度空间积分"尺度"图像,从而得到了新的积分微分方程.这种新的积分微分方程含有一个单调增加的尺度函数.通过选取适当的尺度函数,该方程可以有效地去除乘性噪声.
Multiplicative noise removal is an important research topic on image processing. This paper presents a novel integro-differential equation approach for removing multiplicative noise. Under the assumption that multiplicative noise follows a Gamma distribution, we firstly discuss the classical AA model, and then in order to arrive at the novel integro-differential equation, integrate in inverse scale space a succession of refined ＇slices＇ of the image. The novel integro-differential equation includes a monotone increasing scaling function. By choosing an adaptive scaling function, this equation can remove multiplicative noise efficiently. Finally, the experimental results demonstrate the better performance of the proposed model.
出处
《西安电子科技大学学报》
EI
CAS
CSCD
北大核心
2013年第3期132-138,共7页
Journal of Xidian University
基金
博士点新教师基金资助项目(20100203120010)
国家自然科学基金资助项目(61105011)
关键词
图像去噪
总变差最小
乘性噪声
积分微分方程
image denoising
total variation minimization
multiplicative noise
integro-differential equation
