地理空间信息2024，Vol.22Issue(3)：66-71,6.
基于特色村镇的多源影像最优融合分类方法探讨
Inquire into Optimal Fusion Classification Method of Multi-source Images Based on Characteristic Villages and Towns
摘要
Abstract
The construction of characteristic villages and towns is an important part of development strategy of"Rural Revitalization".Taking He Village,a characteristic village and town in Huzhou City,Zhejiang Province,as the research object,based on multi-source remote sensing images,we used multiple feature variables such as spectrum,texture,geometry and index to design six different feature combination methods for comparing and analyzing multi-source data fusion methods.And then,we selected object-oriented random forest classification method to classify the characteristic villages and towns.The results show that the six multi-source data fusion methods all have high classification accuracy.The classification accuracy of multi-source data fusion is generally better than that of single data.The combination of spatial spectral information of high-resolution data and exponential synthetic image of multi-spectral data can maximize the advantages of multi-source data.关键词
多源数据/特征融合/特色村镇/遥感监测Key words
multi-source data/feature fusion/characteristic villages and towns/remote sensing monitoring分类
天文与地球科学引用本文复制引用
徐聪,于丽君,柳泽,朱建峰,李力,俞燕..基于特色村镇的多源影像最优融合分类方法探讨[J].地理空间信息,2024,22(3):66-71,6.基金项目
"十三五"国家重点研发计划资助项目(2019YFD1100705) （2019YFD1100705）
陕西省自然科学基础研究计划资助项目(2020JQ-742) （2020JQ-742）
西安科技大学培育基金资助项目(6310217017) （6310217017）
西安科技大学博士启动基金资助项目(6310117032). （6310117032）
