高光谱图像主分量变换的一种快速算法

A Fast Approach to Principal Components Transformation for Hyperspectral Image

结合高光谱图像数据的特点,本文从分类和识别的角度,对特征提取在高光谱图像分析中的重要性进行了分析。根据主分量变换的原理,对分组主分量变换方法和常规主分量变换方法的时间花费从理论上进行了对比说明。并且通过对比试验,对文中的方法进行了验证。

From a consideration of efficient classification and recognition of hyperspectral image, this paper analyses the importance of feature extraction. Based on the principle of principal components transformation, this paper makes a theoretical comparison of computational load between conventional method and segmented method. Finally, it validates the efficiency of the proposed method with an experiment.