基于高分辨率遥感影像的不透水面信息提取方法研究

不透水面是城市地区的典型特征,它与城市总用地面积的比值--不透水率作为一个重要的城市生态指数常出现于城市水文、水质、面源污染以及城市植被制图等研究中。利用高分辨率遥感影像提取不透水面不仅可获得较高精度的不透水面信息，而且可为中低分辨率遥感影像的不透水面提取提供样本训练区并检验其提取精度。本文利用南京IKONOS影像，采用面向对象分类方法提取不透水面信息，初步解决了阴影归类和遮盖不透水面的植被剔除等问题，提高了不透水面信息提取精度。

The Study of Extracting Impervious Surface Information Based on High-resolution Remote Sensing Image

Impervious surface is a character of urban areas. The ratio of imperviousness and total area becomes a significant urban ecological index in the research of urban hydrology, water pollution, urban vegetation mapping and so on. Extracting impervious information from high resolution remote sensing image, can not only obtain impervious surface distribution with higher accuracy, but also provide sample training region for the impervious area extraction and accuracy calculation from middle or lower resolution remote sensing image. In this paper, an object oriented classification method to detect impervious surfaces from IKONOS image is proposed. The result shows that the method can partially resolve the problems such as shadows classification and elimination of plants covering impervious area, obtaining a more accurate impervious information.