基于GF-6影像的GNDVI水体指数应用研究

Research on the Application of GNDVI Water Body Index Based on GF-6 Images

基于GF-6影像的光谱分析,本文利用反射率最强的近红外波段和反射率最弱的Coastal Blue波段做差,并与绿波段建立比值关系,结合固定步长搜索方法获取步长改正因子,提出绿波改正差异性归一化植被指数(Green Wave Correction Difference Normalized Vegetation Index,GNDVI)。同其他方法进行对比,研究该指数在水体提取中的应用价值。结果表明:该指数能够有效、稳定地提取水体空间分布。研究区水体提取结果总体精度均在76%以上、F1均在0.68以上、Kappa系数处于高度一致性,为GF-6影像的水体提取提供了一种技术参考。

Based on the spectral analysis of GF-6 images,this article uses the near-infrared band with the strongest reflectance to subtract the Coastal Blue band with the weakest reflectance and establishes a ratio relationship with the green band.Combining with a fixed step search method to obtain the step correction factor,the Green Wave Correction Difference Normalized Vegetation Index(GNDVI)is proposed,which is compared with other methods.The application value of this index in water extraction was studied,and the results showed that it can effectively and stably extract the spatial distribution of water bodies.The overall accuracy of the water extraction results in the research area is above 76%,F1 is above 0.68,and the Kappa coefficient is highly consistent.This provides a technical reference for GF-6 imaging in water extraction.