摘要
各向异性扩散偏微分方程(PDE)的图像平滑方法能够在去噪的同时在一定程度上保持图像的边缘特征,但对图像细节的保持仍不理想。提出采用归一化的3次B样条作为扩散函数,给出了新的PDE模型算法,该算法能在除噪声的同时增强图像边缘细节,并可降低迭代次数,对加性和乘性噪声均有效。
The partial differential equations （PDE） of anisotropic diffusion can be used to suppress the image noises and preserve the edges and detail features of image to some extent, but the denoising effect is still not idea. A new algorithm based on adopting a normalized third order B-spline function as a diffusion one was proposed, which can improve image edge detail besides denoising, decrease iterative order and have an effect on additive and multiplicative noises.
出处
《兵工学报》
EI
CAS
CSCD
北大核心
2008年第8期960-964,共5页
Acta Armamentarii
关键词
信息处理技术
偏微分方程
B样条函数
去噪
边缘保持
information processing technique
partial differential equation
B-spline function
denois-ing
edge preserving
