摘要
为了从遥感影像上对丘陵山区的水田地类快速、准确的提取,以芜湖市南陵县为例,利用5月份的IRS-P6遥感影像和8月份TM遥感影像,分析水田和其背景地物的光谱特征,在ERDAS软件建立基于多源数据的提取模型,对TM影像中的水田地类进行自动提取。结果表明,该方法的总体提取效果较好,与人机交互解译相比,其精度达到92.7%。误判部分可以通过目视解译加以修正。
In order to extract paddy fields in the hilly areas from remote sensing image in a rapid and accurate way, this paper took Nanling County of Wuhu City as an example analyzing spectral features of paddy fields and their background land features with IRS-P6 image of May and TM image of August, built a multivariate data- based extraction model using software ERDAS and then automatically extracted paddy fields from TM image. The results showed that the extraction is good in general and has a precision up to 92.7% compared with man- computer interaction interpretation, and the errors can be corrected by visual interpretation.
出处
《安徽地质》
2015年第4期288-291,共4页
Geology of Anhui
关键词
多时相
丘陵山区
水田
遥感提取
南陵县
multi-temporal
hilly area
paddy field
remote sensing extraction
Nanling County
