摘要
从扩散的角度分析了图像处理中传统热扩散和PM模型的不足,提出了一种各向异性的改进PDE滤波算法.该算法利用图像的结构张量信息和局部特征,自适应选取扩散系数,在图像平坦区域,具有各向同性扩散;而在图像边缘处,则只沿着切向扩散.实验结果表明,该算法具有良好的滤波性能,在滤波的同时可有效保护边缘细节,对灰度图像特别是医学图像,相对于传统方法,该算法可以获得更好的主观视觉效果和客观性能评价指标.
The drawbacks of heat diffusion and PM model are analyzed from the view of diffusion.Then an improved anisotropic filtering algorithm based PDE is proposed.The proposed algorithm can select diffusion coefficients adaptively according to the structure tensor information and local features of the image; in the flat region of the image,diffusion will be executed with equal spread both along the gradient and tangential,while on the edge of the image,diffusion will be executed only along the tangential.The experimental results demonstrate that the proposed algorithm has good filteringperformance,which can preserve edge details effectively during denoising.Meanwhile,for both gray image and medical image,the proposed algorithm is superior to the conventional methods in the aspect of objective performance evaluation and subjective visual effect.
出处
《云南民族大学学报（自然科学版）》
CAS
2015年第1期57-61,共5页
Journal of Yunnan Minzu University:Natural Sciences Edition
基金
四川省教育厅一般科研项目(13ZB0098)
四川文理学院大学生科研项目(X2013Z006)
关键词
图像滤波
偏微分方程
各向异性
扩散系数
结构张量
image filtering
partial differential equations
anisotropy
diffusion coefficient
structure tensor
