摘要
通过CN波谱融合、PCA主成份变换、Brovey变换、HCS变换、HPF高通滤波变换、Pan Sharpen变换、GS光谱锐化、SFIM基于亮度调节的平滑滤波等8种不同的遥感影像融合方法,对北京市延庆县蔡家河流域平原造林地的Pleiades卫星全色和多光谱影像进行融合,通过定性和定量方法从光谱保真、人工造林地林木自动识别精度等方面评价遥感影像融合效果。结果表明:8种影像融合方法各有优势,其中HCS融合方法操作简便,产生的遥感图像光谱保真强,人工造林地林木自动识别精度高,是一种适于Pleiades卫星影像进行林地监测的较好融合方法。
Eight different image fusion methods were adopted,including the CN spectrum fusion,PCA principal component transformation,Brovey transform,HCS transform,HPF high-pass filter transform, PanSharpen transformation,GS spectrum sharpening,based on a brightness control SFIM smoothing,to fuse the Pleiades satellite panchromatic and multi-spectral image of forest in Cai jiahe basin plain affor-estated area,Yanqing county,Beijing.The qualitative and quantitative result of remote sensing image fu-sion on spectral fidelity was analyzed,and the accuracy of automatic identification forest trees in the arti-ficial planting area using fused image was evaluated.Results show that the eight kinds of image fusion methods have their own advantages,but the HCS fusion method is simple,the remote sensing image spectral fidelity,artificial planting area of forests automatic identification precision is highest.So it is a suitable method for the Pleiades satellite images fusion to monitor forest.
出处
《林业资源管理》
北大核心
2016年第3期128-134,共7页
Forest Resources Management
基金
北京市农业科技项目(20140110)