基于NDWI与改进型FCM相结合的高分一号影像水域信息提取方法研究

摘要

鉴于采用传统方法在提取高分辨率影像水域信息存在不足,本文以高分辨率国产卫星高分一号宽幅16 m空间分辨率多光谱影像为数据源,以浙江省钱塘江流域为研究区域,通过预先对影像进行NDWI处理,在此基础上利用改进FCM聚类算法实现了水域信息的自动提取,并与传统的NDWI和ISODATA分类方法提取结果进行比较.结果表明,该方法在分类效果和精度方面均优于传统分类方法,作为国产高分一号影像水域信息自动提取方法是可行的.%In view of the shortcomings of traditional methods in extracting water information in high resolution images,a modified FCM clustering algorithm after being proposed by NDWI in advance was used to extract the information of water body.The China-made GF -1 muhispectral imagery with 16m spatial resolution was taken as the data source and the district of Qiantang Rive in Zhejiang Province was taken as the study area in this paper.The result was compared with the traditional NDWI and ISODATA classification methods.The results show that the method proposed in this paper is superior to the traditional methods in classification effect and precision,and it is feasible as an automatic water body extraction method for China-made GF-1 image.