基于非子采样金字塔变换的高光谱图像异常检测  

Anomaly Detection for Hyperspectral Image Based on Nonsubsampled Pyramid Transform

作　　者：宗靖国[1] 张建奇[1] 秦翰林[1] 刘德连[1] 

机构地区：[1]西安电子科技大学技术物理学院,陕西西安710071

出　　处：《红外技术》2011年第1期56-60,共5页Infrared Technology

基　　金：国家自然科学基金资助项目(编号:60777042)

摘　　要：针对高光谱图像异常检测中背景问题,提出了一种新的基于非子采样金字塔变换的异常检测方法。首先对高光谱图像进行分解,将其划分为不同频率的子块;根据子块中异常和背景的特点构造变分偏微分方程来抑制背景杂波;进而将检测算子应用于抑制后的子块;重构图像,最终得到异常检测的结果。采用真实和模拟的超光谱图像异常点进行了仿真实验,与经典的RX算法相比,实验结果验证了该算法具有很好的检测性能。A new anomaly detection method for hyperspectral images based on Nonsubsampled Pyramid Transform was presented to solve the anomaly detection of hyperspectral images .Firstly the hyperspectral images is decomposed into a series of different frequency sub-bands. Then according to the feature of sub-bands, the variation of partial differential equation is proposed to suppress background clutter. Furthermore the detection operator is used in the sub-bands and the detection result is obtained by reconstructing the sub-bands. The real and simulation hyperspectral image anomaly targets were used to test. Comparing with RX algorithm, the experimental results have proved the validity of the new method.

关 键 词：高光谱 异常检测 非子采样金字塔变换 变分偏微分方程 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]