基于随机森林算法的珠海一号高光谱影像土地利用信息提取  

Land use information extraction from hyperspectral images of zhuhai-1 based on random forest algorithm

作　　者：杨远陶 刘瑞[1,2] 曹礼刚[1,2] 杨梅 陈景珏[3] YANG Yuantao;LIU Rui;CAO Ligang;YANG Mei;CHEN Jingjue(Chengdu University of Technology College of Geophysics,Chengdu 610059,China;Key Lab of Earth Exploration&Infomation Techniques of Ministry Education,Chengdu 610059,China;College of Earth Sciences,Chengdu 610059,China)

机构地区：[1]成都理工大学地球物理学院,成都610059 [2]成都理工大学地球勘探与信息技术教育部重点实验室,成都610059 [3]成都理工大学地球科学学院,成都610059

出　　处：《物探化探计算技术》2021年第6期818-824,共7页Computing Techniques For Geophysical and Geochemical Exploration

基　　金：中国地质调查局项目(20180606-01)。

摘　　要：以珠海一号高光谱卫星遥感影像作为实验数据,提出了一种结合波段标准差和随机森林算法的高光谱遥感图像分类方法。首先通过利用波段的标准差筛选出合适的波段,然后选取适量的样本,作为模型训练和验证的样本。随后将随机森林模型加入训练样本进行模型训练,将训练好的模型用于原始影像预测得到最终的分类结果。最后将分类结果结合验证样本评价随机森林模型的分类精度。结果表明,利用随机森林算法的分类总精度达到了93.14%,而最大似然法分类总精度为86.38%,分类总精度提高了6.76%,表明随机森林算法在高光谱影像提取土地利用信息中有明显的优势,可为高光谱影像在土地利用信息提取中提供一种可行的方法。In this paper,a hyperspectral remote sensing image classification method that combines band standard deviation and random forest algorithm was proposed by adopting the Zhuhai-1 hyperspectral satellite images.First,we chose some appropriate bands based on the standard deviation of the bands,and then an appropriate amount of samples were selected and divided into training samples and validation samples.The training samples were subsequently trained by the random forest algorithm,and the final classification results were obtained after the addition of the original images.Finally,the validation samples were implemented to evaluate the accuracy of the classification results.The results show that the total classification accuracy of the random forest algorithm reaches 90.85%,while the total classification accuracy of the maximum likelihood method is 82.35%.The accuracy of the former is 8.5%higher than that of the latter,indicating that the random forest algorithm has obvious advantages,and it is also an efficient method in land use information extraction from hyperspectral images.

关 键 词：随机森林 珠海一号 高光谱 土地利用 

分 类 号：P631.4[天文地球—地质矿产勘探]