| 引用本文: | 粘永健,王展,万建伟,等.面向异常检测的高光谱图像压缩技术.[J].国防科技大学学报,2009,31(3):48-52.[点击复制] |
| NIAN Yongjian,WANG Zhan,WAN Jianwei,et al.Compression Technique for Hyperspectral Imagery Oriented Anomaly Detection[J].Journal of National University of Defense Technology,2009,31(3):48-52[点击复制] |
|
| |
|
|
| 本文已被：浏览 7485次 下载 5705次 |
| 面向异常检测的高光谱图像压缩技术 |
| 粘永健, 王展, 万建伟, 辛勤 |
|
(国防科技大学 电子科学与工程学院，湖南 长沙 410073)
|
| 摘要: |
| 异常检测已经成为高光谱图像重要的后续应用之一，提出了一种面向异常检测的高光谱图像压缩算法。为减少压缩对异常检测性能的影响，首先采用RX算子对高光谱图像进行异常检测，并对异常点与背景对应的光谱维矢量进行预处理。对高光谱图像的光谱维矢量进行KL变换，通过引入虚拟维数估计算法对原始数据的本征维数进行估计，在此基础上给出了一种主分量选取方法。最后，采用最优码率分配策略为各主分量分配相应的压缩码率，并利用SPIHT算法分别进行压缩。实验结果表明，该算法在获得较高压缩性能的同时，可有效保持图像中的异常信息。 |
| 关键词: 高光谱图像 有损压缩，异常检测 虚拟维数 |
| DOI： |
| 投稿日期：2009-02-12 |
| 基金项目:国家自然科学基金资助项目(60572135) |
|
| Compression Technique for Hyperspectral Imagery Oriented Anomaly Detection |
| NIAN Yongjian, WANG Zhan, WAN Jianwei, XIN Qin |
|
(College of Electronic Science and Engineering, National Univ. of Defense Technology, Changsha 410073, China)
|
| Abstract: |
| Anomaly detection has been one of the most important applications for hyperspectral imagery. A new lossy compression method for hyperspectral imagery oriented anomaly detection is proposed. In order to keep the performance of anomaly detection, the anomalous vectors detected by the improved RX algorithm are preprocessed. Furthermore, virtual dimensionality algorithm is introduced to estimate the Intrinsic Dimensionality (ID) of original data while Karhunen-Loeve transform is used to provide spectral decorrelation. In addition, based on the virtual dimensionality, a new method for the number of principle component determination is presented. The bit rate of each principle component is distributed by optimal rate allocation strategy for spatial compression by SPIHT algorithm. Experimental results show that the proposed algorithm provides better compression performance, as well as efficient preservation for anomalous pixels. |
| Keywords: hyperspectral imagery lossy compression anomalous detection virtual dimensionality |
|
|
|
|
|
