论文部分内容阅读
应用融合变换对四川西昌地区紫茎泽兰遥感图像ASTER数据进行处理,结果表明:对ASTER三种传感器生成的三种分辨率的影像数据进行Gram-Schmidt融合处理,可以将具有较高光谱分辨率的多光谱遥感波段的空间分辨率从30m×30m提高到15m×15m,增加更多的地表地物组分信息,有助于识别各种不同地物类型,同时使同一地物的光谱信息变换前后保持不变,确保了后续图像分类的可靠性。应用ENVI中对融合后的图像进行马氏距离分类,分类结果总体精度为73.6983%,Kappa系数等于0.6936。
The ASTER data of Eupatorium adenophorum in Xichang, Sichuan Province were processed by using the fusion transform. The results showed that Gram-Schmidt fusion of the three kinds of data generated by the ASTER sensors could be used to transform the ASTER data with higher spectral resolution , The spatial resolution of the multi-spectral remote sensing band is increased from 30m × 30m to 15m × 15m, and more surface feature components are added to help identify different types of landforms and transform the spectral information of the same landform Before and after the same, to ensure the reliability of follow-up image classification. Applying Mahalanobis distance to the fused images in ENVI, the overall accuracy of classification results is 73.6983%, and the Kappa coefficient is equal to 0.6936.