Tree Species Classification Based on Fusion Images by GF-5 and Sentinel-2A

Forest ecosystem detection and assessment usually requires accurate spatial distribution information of forest tree species. Remote sensing technology has been confirmed as the most important method for species acquisition, space-borne hyperspectral imagery, with advantages high spectral resolution, provides a better possibility classification. However, present in-orbit imager proved to be too low in resolution meet accuracy needs In this study, we firstly explored evaluated effectiveness Gram-Schmidt (GS) Harmonic analysis fusion (HAF) image GaoFen-5 (GF-5) Sentinel-2A. Then, Integrated Z-Score (IFZ) was used extract from fused image. Next, textural features image, topographic extracted DEM were selected according random importance ranking (Mean Decreasing Gini (MDG) Mean Accuracy (MDA)), imported into classifier complete The results showed that: comparing some evaluation factors such entropy, average gradient standard deviation images, GS proven have higher degree integration fidelity. that WBI, Aspect, NDNI, ARI2, FRI more Both classification kappa coefficients images significantly greatly improved when compared those original GF-5 images. overall ranged 61.17% 86.93% different feature combination scenarios, based on MDA achieved (OA = 86.93%, Kappa 0.85). This study demonstrated feasibility Sentinel-2A classification, which further good reference application