基于高光谱图像主成分分量的小目标检测算法研究

提出了一种基于图像主成分分量的高光谱小目标检测算法．作为一种多元数据集合，通常高光谱数据形成的几何体是一个超平面．主成分分析能有效估计这一几何体的本征维数．显著特征值对应的主成分体现了几何体大部分信息；而不显著特征值对应的主成分则代表了正交于几何体的信息，而这些信息中则包含了重要的内容，例如目标特性。文中提出的方法就是利用这些不显著的主成分分量来进行小目标检测．该方法减少了对先验光谱信息的依赖，提高了算法的实用性．

ALGORITHM ON SMALL TARGET DETECTION BASE ON PRINCIPAL COMPONENT OF HYPERSPECTRAL IMAGERY

A small target detection approach which is based on principal component of hyperspectral imagery was presented. As a kind of multivariate data, the points of hyperspectral data always compose a hyper-plane in high-dimensional space. The principal component analysis can estimate the intrinsic dimensionality of the hyper-plane.Usually, the significant components contain most information of imagery. The insignificant components, which covers important details such as target characteristic, represent the information of orthogonal subspace. These insignificant principle components were used to detect small targets. This method reduces the dependent of the spectral pre-information and improves the practicability.