融合变换在紫茎泽兰遥感监测中的应用研究

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。

Study on Application of Fusion Transformation in Remote Sensing-monitoring of Eupatorium Adenophorum.

In this article, the remote sensing data was treated by fusion transformation, which down load from ASTER of Eupatorium adenophorum in Xichang district Siehuan province. The results showed that there were three image datas of different resolution, which came from the three kinds of sensors of ASTER. The data was treated by the Gram-Schmidt fusion processing, which enhanced the spatial resolution of muhi-spectrum remote sensing wave band with high spectrum resolution from 30m×30m to 15m×15m. The treat obtained more information of land surface and ground objects feature component, which was helpful to distinguish kinds of terrain feature; and kept the treated spectrum information same as before, which ensured the reliability of following image classification. The fusion images were classified through Mahalanobis Distance Classifier of ENVI, which overall precision was 73.6983% , the coefficient of Kappa was 0.6936.