Spectral Estimation of Carotenoid Density in Populus pruinosa Leaves

Forests, Journal Year: 2025, Volume and Issue: 16(3), P. 394 - 394

Published: Feb. 23, 2025

Carotenoids play a crucial role in the photosynthesis process plants. Estimating and modeling carotenoid content Populus pruinosa leaves via high-spectrum technology is highly important for health status monitoring. This study involved acquiring spectral reflectance of at different times, followed by smoothing data with Savitzky—Golay filter, then using methods such as first derivative (FD), continuous wavelet transform (CWT), first-order combined (CWT+FD), creating three transformation methods. Two- three-dimensional vegetation indices were constructed unified manner. Two methods, backpropagation neural network (BPNN) support vector regression (SVR), employed to estimate leaf density combining indices. The results show that after canopy processed FD, CWT, CWT+FD on basis SG smoothing, it can effectively highlight characteristics leaves, local absorption features are more significant. Compared preprocessing showed correlation between values FD + CWT method highest. three-band index exhibited 4.26% stronger than did two-band index. Among index-based models, SVR model outperforms BPNN model. For chlorophyll density, based achieves best performance. coefficient determination (R2) set was 0.782, root-mean-square error (RMSE) 0.022, relative percentage deviation (RPD) 0.206. validation set, value 0.648, RMSE 0.023, RPD 1.526, indicating accuracy.

Language: Английский

Monitoring and Risk Prediction of Low-Temperature Stress in Strawberries through Fusion of Multisource Phenotypic Spatial Variability Features

Plant Phenomics, Journal Year: 2025, Volume and Issue: unknown, P. 100041 - 100041

Published: May 1, 2025

Language: Английский