摘要
以芜湖市为例,选用2007年成像的TM影像数据,采用阈值分类与整体—局部多尺度迭代模型相结合的方法,对该地区绿地信息进行提取,并与监督分类方法、阈值分类法提取的精度进行比较。结果表明,阈值分类与整体—局部多尺度迭代模型相结合的方法明显高于监督分类法和阈值分类法,在适宜阈值取165时,其用户精度和生产精度分别达84.21%和87.89%。
A study was conducted to extract the remote sensing information of greenbelt in Wuhu City based on TM image data in 2007 using a combined method of threshold classification and multi-scale iterative model.The extraction result was compared with both methods of supervised classification and threshold classification.Results show that the combined method is superior to the methods of supervised classification and threshold classification.When the threshold was set to 165,the user＇s accuracy and the producer＇s accuracy reached 84.21% and 87.89% respectively.
出处
《东北林业大学学报》
CAS
CSCD
北大核心
2010年第12期77-79,共3页
Journal of Northeast Forestry University
关键词
绿地信息
遥感
芜湖市
多尺度
Greenbelt information
Remote sensing
Wuhu city
Multi-scale
