高光谱影像目标探测可视为一个分类问题,本文通过揭示支持向量回归(SVR)与支持向量分类(SVC)之间的关系,证明了SVR用于分类的可行性,并以此为根据提出了一种基于SVR的目标探测算法,该算法利用虚拟维数得到端元个数的估计,结合端元选择和线性混合模型生成训练样本替代从影像中选择的训练样本,因而减少了对影像先验知识的依赖。采用模拟数据和由AVIRIS获得的高光谱影像对本文算法进行了检验,结果令人满意。 Hyperspectral image target detection can be considered as a classification problem. This paper proves the feasibility of using SVR for classification by revealing the relationship between support vector regression (SVR) and support vector classification (SVC), and based on this, A target detection algorithm based on SVR, which uses the virtual dimension to get the estimation of the number of end-points, combines the end-element selection and the linear mixed-model generation to generate the training samples instead of the training samples selected from the images, The dependence of knowledge. The proposed algorithm has been tested with simulated data and hyperspectral images obtained from AVIRIS with satisfactory results.