摘要
均值曲率运动(MCM)方程能有效去除脉冲噪声。给出了MCM方程的一种显示差分格式。针对ENoD-8格式在去除高密度噪声时会残留斑块的问题,提出了新的ENoD-A格式。该格式的应用减少了迭代过程中的计算量,提高了去噪效率。实验结果表明,新格式增强了方程去除椒盐噪声的能力,同时又保护了图像细节。
Mean Curvature Motion(MCM) equation can effectively eliminate the impulse noise.This paper presents an explicit difference scheme for MCM equation,and a new scheme ENoD-A is proposed by analyzing the shortcomings of ENOD-8. The application of this scheme reduces the computational complexity in the iterative process and improves the efficiency of the denoising.Numerical experiments have shown that the new scheme enhances the ability of equations to eliminate impulse noise,while protecting the image texture detail.
出处
《计算机工程与应用》
CSCD
北大核心
2010年第25期164-167,198,共5页
Computer Engineering and Applications
基金
云南省教育厅科学研究基金项目No.09SKJ03~~
关键词
图像去噪
均值曲率运动(MCM)
椒盐噪声
ENoD差分格式
image denoising
Mean Curvature Motion(MCM)
impulse noise
Essentially Non-Dissipative(ENoD) difference scheme