基于Sentinel-2A影像的互花米草信息提取研究

Research on Information Extraction of Spartina Alterniflora Based on Sentinel-2A Image

利用面向对象-支持向量机分类方法,基于2019年Sentinel-2A影像进行互花米草提取,并通过最优尺度估计确定最优分割参数,继而实现多尺度分割以及分类特征构建等对影像进行分类。研究结果表明:面向对象-支持向量机分类方法能够很好地将互花米草区分出来,分类的总体精度达到89%,Kappa系数达到86%。另外,2019年九段沙互花米草的分类面积达到5856.85 hm^(2),并主要分布在中沙以及下沙中上潮间带,在上沙有小部分分布。

Using the object-oriented-support vector machine classification method to remotely the 2019 Sentinel-2A images were interpreted,and the optimal segmentation parameters were determined through optimal scale estimation,and then the images were classified through multi-scale segmentation and construction of classification features.The research results show that the object-oriented-support vector machine classification method can distinguish spartina alterniflora well,with an overall accuracy of 89%and a Kappa coefficient of 84%.In addition,in 2019,the classified area of spartina alterniflora in Jiuduansha reaches 5856.85 hm^(2),and it is mainly distributed in Zhongsha and Xiasha intertidal zones,and a small part of it is distributed in Shangsha zone.