| 基于各向异性扩散的几种平滑算法比较及改进 |
| |
| 引用本文： | 钱惠敏,茅耀斌,王执铨.基于各向异性扩散的几种平滑算法比较及改进[J].南京理工大学学报(自然科学版),2007,31(5):605-611. |
| |
| 作者姓名： | 钱惠敏 茅耀斌 王执铨 |
| |
| 作者单位： | 南京理工大学,自动化学院,江苏,南京,210094 |
| |
| 摘 要： | 该文比较了几种现有的P-M扩散系数改进算法的优缺点,提出了一种改进的基于视觉掩蔽效应的各向异性扩散平滑算法.该方法用归一化局部方差和图像梯度信息联合度量空间细节,能更有效地消除噪声,保持图像细节,也避免了原扩散系数参数估计的困难.实验结果表明,提出的改进方法得到的平滑图像具有更好的视觉质量.
|
| 关 键 词： | 各向异性扩散 视觉掩盖效应 图像平滑 各向异性扩散 平滑图像 算法比较 改进方法 Anisotropic Diffusion Based Algorithms Smoothing Improvement 视觉质量 结果 实验 参数估计 图像细节 消除噪声 度量空间 梯度信息 方差 局部 归一化 |
| 文章编号： | 1005-9830（2007）05-0605-07 |
| 修稿时间： | 2006-03-21 |
Comparison and Improvement of Several Smoothing Algorithms Based on Anisotropic Diffusion |
| |
| QIAN Hui-min,MAO Yao-bin,WANG Zhi-quan.Comparison and Improvement of Several Smoothing Algorithms Based on Anisotropic Diffusion[J].Journal of Nanjing University of Science and Technology(Nature Science),2007,31(5):605-611. |
| |
| Authors: | QIAN Hui-min MAO Yao-bin WANG Zhi-quan |
| |
| Institution: | School of Automation, NUST, Nanjing 210094, China |
| |
| Abstract: | The performances of several improved Perona-Malik anisotropic diffusion methods are compared, and an improved method making use of visual masking effect is proposed. Owing to the usage of the measurement capability for image features of unitary local variance and image gradient, the new algorithm achieves better effect in terms of noise removal and edge preservation and obviates the parameter estimation difficulties that unavoidably exist in old schemes. The experimental results demonstrate the better smoothing images from the algorithm with respect to visual quality against the many existing improved anisotropic diffusion methods. |
| |
| Keywords: | anisotropic diffusion visual masking effect image smoothing |
| 本文献已被 维普 万方数据 等数据库收录！ |
