摘要
本文在遥感技术的支持下将崇明岛东滩作为研究区域。其间使用2017年2—10月Sentinel-1A微波影像和2017年2—9月的Landsat8和sentinel-2A的光学影像。预处理后,通过计算,使用决策树分类对崇明东滩的3种重要植被(互花米草、芦苇、海三菱藨草)进行分类,得到多时相条件下不同植被的分类结果,用混淆矩阵做精度评价。研究表明,光学与微波相结合的多时相分类图有更好的精度。
With the support of remote sensing technology,this paper took Dongtan of Chongming Island as the re⁃search area.In the meantime,Sentinel-1A microwave images from February to October 2017 and Landsat8 and senti⁃nel-2A optical images from February to September 2017 was used.After pre-processing,the decision tree classifica⁃tion was used to classify the three important vegetations(Spartina alterniflora,Phragmites australis,and Sea Mitsubi⁃shi)in Chongming Dongtan through calculation and obtain the classification results of different vegetations under multi-temporal conditions,and confusion matrix was used for accuracy evaluation.Research shows that the multitemporal classification map combined with optics and microwave has better accuracy.
作者
王明昊
白万晟
刘松涛
WANG Minghao;BAI Wansheng;LIU Songtao(Chengdu University of Information Technology,Chengdu Sichuan 610225)
出处
《河南科技》
2020年第26期148-150,共3页
Henan Science and Technology
关键词
湿地
崇明岛东滩
光学
微波
植被分类
wetland
Chongming Island Dongtan
optics
microwave
vegetation classification
