一种新的高光谱图像分类方法

摘要

基于像元波谱曲线特征提取进而利用所提取的特征进行分类是高光谱图像分类的重要研究内容。提出了一种基于傅里叶变换幅度谱的高光谱遥感图像分类新方法，也就是利用像元波谱曲线傅里叶变换幅度谱最值单一特征实现对地物的分类识别。%Based on the feature pixel spectrum curve extraction and use the extracted features for classification is an important research content of hyper-spectral image classification. Puts forward a new method of hyper-spectral remote sensing image classification based on Fourier transform spectrum, namely the use of pixel spectral curves of Fourier transform spectrum maximum single feature classification and recognition of objects.