基于光谱和纹理特征的山区高分辨率遥感影像分类

Land use/cover classification of high resolution remote sensing images of mountainous areas based on texture and spectral information

本文在只做阴影补偿而不做地形校正的情况下,使用光谱和纹理特征相结合的方法进行山区高分辨率遥感影像分类。实验取得了78%的分类精度,表明该方法合理可行,具有一定的实用性。

In this paper, only the shadow compensation has been done, without the topographic correction. Using the method integrating the spectral and texture features to classify land use/cover types by high-resolution remote sensing image from mountainous region. Experiment has been made and achieved a classification accuracy of 78%, indicating that the method is reasonable and practicable.