Combining Different Transformations of Ground Hyperspectral Data with Unmanned Aerial Vehicle (UAV) Images for Anthocyanin Estimation in Tree Peony Leaves

To explore rapid anthocyanin (Anth) detection technology based on remote sensing (RS) in tree peony leaves, we considered 30 species of peonies located Shaanxi Province, China. We used an SVC HR~1024i portable ground object spectrometer and mini-unmanned aerial vehicle (UAV)-borne RS systems to obtain hyperspectral (HS) reflectance images canopy leaves. First, performed principal component analysis (PCA), first-order differential (FD), continuum removal (CR) transformations the original ground-based spectra; commonly spectral parameters were implemented estimate Anth content using multiple stepwise regression (MSR), partial least squares (PLS), back-propagation neural network (BPNN), random forest (RF) models. The transformation highlighted characteristics curves improved relationship between Anth, RF model FD spectrum portrayed best estimation accuracy (R2c = 0.91; R2v 0.51). Then, RGB (red-green-blue) gray vegetation index (VI) texture constructed UAV images, was parameters. Finally, image fused with data, a multisource constructed, PCA + UAV, CR MSR, PLS, BPNN, methods. FD+UAV modeling verification effect 0.93; 0.76); compared FD-RF model, R2c increased only slightly, but greatly from 0.51 0.76, indicating testing accuracy. optimal for leaves obtained, high-precision constructed. Our results can be selection HS future plant estimation, as theoretical basis growth monitoring RS.