摘要
以Landsat-8为数据源,美国波士顿为研究区,在归一化差异不透水指数(NDISI)基础上,进行热红外波段细化,提出修正归一化差异不透水面指数(Modified NDISI,MNDISI),并对二者进行比较。结果表明,NDISI得到的不透水面总体精度和OK系数分别为84.19%和0.68,MNDISI提取的不透水面总体精度和OK系数达到87.52%和0.749。MNDISI相比NDISI,提取不透水面的效果更优。
In order to improve the resolution of thermal band, a modified normalized difference impervious surface index was proposed to extract urban impervious surface from Landsat 8 image in this paper. In order to evaluate the performance of MNDISI for mapping urban impervious surface, Boston City of United States was selected as study area, and comparison analysis was conducted between MNDISI and normalized difference impervious surface index(NDISI). The results show that MNDISI can enhance the urban information, and is an effective index of impervious surface extraction with more than 87% corresponding to overall accuracy(OA). Therefore, our proposed MNDISI is proved to be an effective impervious surface index for mapping ISA distribution with high accuracy at regional and global scale.
出处
《地理空间信息》
2018年第2期90-94,共5页
Geospatial Information
基金
国家重点研发计划资助项目(2016YFA0600302)
中国科学院国际合作局对外合作重点资助项目(131C11KYSB20160061)
国家自然科学基金资助项目(41201357)
