| 噪声抑制的高光谱图像虚拟维数分析 |
| |
| 引用本文： | 何秉钧,蒋鸣飞,罗欣,王蓉.噪声抑制的高光谱图像虚拟维数分析[J].四川大学学报(自然科学版),2017,54(2):303-308. |
| |
| 作者姓名： | 何秉钧 蒋鸣飞 罗欣 王蓉 |
| |
| 作者单位： | 四川农业大学,中国电子科技集团第十研究所,电子科技大学,电子科技大学 |
| |
| 基金项目： | 国家973计划项目（2013CB733400）；中央高校基本科研业务费项目(ZYGX2013J120)；中国博士后科学基金项目（20100471664）；电子科技大学本科教育教学研究项目（2015XJYYB088）。 |
| |
| 摘 要： | 在高光谱数据降维过程中,通常用虚拟维数来表征数据的本征维数.经典的虚拟维数分析算法主要运用假设检验准则设定特征值门限,通过特征值判定来决定虚拟维数值.但是,在强噪声干扰下,经典算法不能有效分析出虚拟维数值.本文提出了一种噪声抑制的高光谱图像虚拟维数分析方法(NCVD),该算法通过对数据矩阵进行QR分解,减小了算法的运算量;采用滑动噪声检测窗口对噪声成分进行滤除,提高了估计维数的准确性;结合最小二乘算法对判别门限进行修正,使虚拟维数估计结果更具合理性;采用模拟数据和真实数据进行实验,实验结果证明,本文所提算法的可行性和较现有算法的优越性.
|
| 关 键 词： | 高光谱图像 虚拟维数 QR分解 滑动噪声检测 |
| 收稿时间： | 2016/7/26 0:00:00 |
| 修稿时间： | 2016/8/28 0:00:00 |
Virtual Dimensionality Analysis of Hyperspectral Imagery with Noise being Constrained |
| |
| HE Bin-Jun,JIANG Ming-Fei,LUO Xin and WANG Rong.Virtual Dimensionality Analysis of Hyperspectral Imagery with Noise being Constrained[J].Journal of Sichuan University (Natural Science Edition),2017,54(2):303-308. |
| |
| Authors: | HE Bin-Jun JIANG Ming-Fei LUO Xin and WANG Rong |
| |
| Institution: | Sichuan Agricultural University,The 10th Research Institute of Chinese Electronic Technology Group Corporation,University of Electronic Science and Technology of China and University of Electronic Science and Technology of China |
| |
| Abstract: | In dimensionality reduction process of hyperspectral data, intrinsic dimension is normally characterized by virtual dimension. Classic algorithm mainly uses hypothesis-testing criterion to set eigenvalue threshold and correspondingly obtains virtual dimension. But under strong noises, it may not estimate very well. A noise constrained virtual dimension (NCVD) analysis method of hyperspectral imagery is proposed in this paper. It decreases the computational complexity by the QR decomposing; improves the accuracy of the estimated dimension by adopting sliding noise detection window to filter the noise; synthesizes the least squares algorithm to modify threshold for reasonable results. The experimental results prove the feasibility and superiority of the proposed algorithm by using simulated and real data. |
| |
| Keywords: | hyperspectral imagery virtual dimensionality QR decomposition sliding noise detection |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《四川大学学报(自然科学版)》浏览原始摘要信息 |
| 点击此处可从《四川大学学报(自然科学版)》下载免费的PDF全文 |
