河南城建学院学报2025，Vol.34Issue(2)：80-89,113,11.DOI:10.14140/j.cnki.hncjxb.2025.02.011
叠加特征变量的GF-6 WFV遥感影像土地利用变化研究
Land use change in GF-6 WFV remote sensing images with superimposed feature variables
摘要
Abstract
Clarifying the status and future development of urban land use is helpful for national land space planning and urban space resource allocation.In this study,GF-6 WFV remote sensing data was used to con-struct four feature variables,namely NDVI,EVI,BDI and NDWI,and random forest classification method was used to explore the spatial-temporal distribution characteristics of land use in Pingdingshan City from 2019 to 2023.The results show that for the random forest classification method,the overall classification accuracy of point training samples(94.84％)is significantly higher than that of surface training samples(88.35％),and the detail expression ability of classification results is also more capable.The superimposed feature variable da-ta can significantly reduce the misclassification error and missing error,and the overall classification accuracy increases from 97.65％to 97.85％,therefore the final ground object classification effect is better than the o-riginal GF-6 WFV data.There is no significant land use change in Pingdingshan from 2019 to 2023,but affect-ed by the policy of"returning farmland to forest"in Henan Province,the area of urban construction gradually decreased,while the area of forest land gradually increased.关键词
GF-6 WFV/特征变量/随机森林/平顶山市/土地利用Key words
GF-6 WFV/feature variables/random forest method/Pingdingshan City/land use分类
测绘与仪器引用本文复制引用
张志敏,张森林,鲁春阳,王丽美..叠加特征变量的GF-6 WFV遥感影像土地利用变化研究[J].河南城建学院学报,2025,34(2):80-89,113,11.基金项目
河南省哲学社会科学规划项目(2022BJJ026) （2022BJJ026）
河南省科技攻关项目(242102320345,242102321123) （242102320345,242102321123）
河南省软科学项目(242400410624) （242400410624）
河南城建学院青年骨干教师培养项目(YCJQNGGJS202307) （YCJQNGGJS202307）
河南城建学院科教融汇项目(K-X2025083) （K-X2025083）
