高光谱遥感图像异常检测研究进展

高光谱遥感图像异常检测即是对遥感图像中与背景存在显著光谱差异的像元进行识别,它在灾害预测和环境监测等领域具有很大应用价值.异常检测受到遥感学界的普遍重视,这些年来对异常检测开展的研究取得了不少成果.高光谱图像异常检测研究最初基于传统数据处理和信号分析方法,随着高光谱图像数据处理研究的深入及数据挖掘技术和智能学习方法等相关技术的发展,现在呈现智能和联合处理的新方向.文章对异常检测算法近来的研究进展进行的介绍以期为异常检测的相关研究提供一定参考.

Research Advance of Anomaly Detection for Hyperspectral Imagery

Anomaly detection of hyperspectral remote sensing image is to identify the pixels with significant different spectral signatures from the neighboring background, it has great application value in fields such as disaster prediction and environment monitoring. Anomaly detection has been taken seriously by the remote sensing community, the related scholars carried out lots of research and yielded many results. At first, anomaly detection was based on traditional data processing and signal analysis methods. Along with the advancement of research in hyspectral image data processing and the development of data mining technology and intelligent learning methods, intelligent and combined treatment occurs at present. By introducing the research advance of anomaly detection algorithms, this paper aims to provide a reference to relevant research on anomaly detection.