Published in Scientific Papers. Series A. Agronomy, Vol. LXVI, Issue 2
Written by Ciprian BUZNA, Marinel Nicolae HORABLAGA, Florin SALA
The study used the technique based on aerial images (UAV) to estimate the production of the sunflower crop, hybrid P64LE25. The experiment was organized within SCDA Lovrin, Timis County, Romania, in the 2021-2022 agricultural year, under the conditions of a chernozem type soil, non-irrigated system. The images were taken with the drone (DJI Phantom 4), at variable heights (H, 3 to 50 m). From the analysis of the digital images, the values of the RGB color parameters were obtained. The distribution of the RGB data series was of normal type (r = 0.950 for R, r = 0.960 for G and r = 0.965 for B). Very strong correlations were recorded between color parameters (r = 0.994 for R and G, r = 0.954 for R and B, r = 0.948 for G and B), and weak correlations between RGB and H. From PCA, PC1 explained 97.698% of variance, and PC2 explained 2.1027% of variance. Regression analysis facilitated the prediction of sunflower production based on RGB values, under statistical safety conditions (p<0.001; RMSEP = 1.7907 in relatin to R and G; RMSEP = 4.5869 in relation to R and B; RMSEP = 3.6978 in relation to G and B).
[Read full article] [Citation]
