摘要
针对城市绿地信息提取时不同影像融合方法缺少系统性研究的问题,本文以国产高分一号、WorldView-2卫星影像的数据为例,借助GS变换、PCA变换以及HIS变换3种方法,融合高分辨率卫星影像,并从融合质量及城市绿地信息提取精度等方面评价了影像融合方法的可行性。研究结果表明,在3种影像融合方法中,GS融合的效果最好;且用于城市绿地信息提取时,GS和PCA能够获得高精度的卫星影像,对影像分类精度的提高效果最为明显,充分满足了城市绿地信息提取的应用需求。
Aiming at the lack of systematic research on different image fusion methods for urban green space information extraction,this paper takes the data of domestic Gaofen-1,WorldView-2 and other satellite images as an example,and fuses high-resolution satellite images with the help of three methods,such as GS transform,PCA transform,and HIS transform,and evaluates the feasibility of the image fusion methods in terms of fusion quality and accuracy of urban green space information extraction.The feasibility of the image fusion methods was evaluated in terms of fusion quality and accuracy of urban green space information extraction.The results show that among the three image fusion methods,GS fusion has the best effect,and when used for urban green space information extraction,GS and PCA can obtain high-precision satellite images,which has the most obvious effect on the improvement of image classification accuracy,and fully meets the application requirements of urban green space information extraction.
作者
杨阿兰
YANG A'lan(School of Architecture Engineering,Sanmenxia Polytechnic,Sanmenxia,Henan 472000,China)
出处
《自动化应用》
2023年第14期186-188,193,共4页
Automation Application
基金
三门峡市2020年科技发展计划项目(2020040311)。
关键词
高分辨率
卫星影像
城市绿地信息提取
影像融合
精度评价
high resolution
satellite image
urban green space information extraction
image fusion
accuracy evaluation
