摘要
高斯滤波算法在去噪时能平滑图像,但是会破坏图像的边缘细节,而基于PDE的各向异性扩散的P&M模型算法在去噪时能保留图像的边缘细节,但是会出现零散的斑点。结合两种算法的优点,通过对扩散系数进行改进,提出一种改进型P&M模型算法。仿真结果表明,该算法能够有效地去除噪声图像中的高斯白噪声和椒盐噪声,能够更好地保留图像的边缘细节,与高斯滤波算法和P&M模型算法相比,改进型P&M模型算法具有更好的去噪性能。
Gaussian filtering algorithm can smooth image when it is used to denoise image, but it will destroy the edge details of image. Anisotropic diffusion P & M model algorithm based on PDE can retain the edge details of image when it is used to denoise image, but scattered dots will appear. Combined with the advantages of two algorithms, a modified P & M model algorithm is proposed by improving diffusion coefficient. The simulation results show that the proposed algorithm can remove Gaussian white noise and salt and pepper noise in noise image effectively and can retain the edge details of image better. The denoising performance of modified P & M model algorithm is better than that of Gaussian filtering algorithm and P & M model algorithm.
出处
《长春工业大学学报》
CAS
2009年第3期301-306,共6页
Journal of Changchun University of Technology
关键词
图像去噪
偏微分方程
高斯滤波算法
P&M模型算法
扩散系数
各向异性扩散
梯度阈值
image denoising
partial differential equation
Gaussian filtering algorithm
P & M modelalgorithm
diffusion coefficient
anisotropie diffusion
gradient threshold.