Validation Testing of UAV-based Vegetable Growth Estimation Models

Abstract. On-site monitoring of biophysical parameters, such as a leaf length, a leaf area, and a fresh weight, in an agricultural field can provide useful information for farmers to establish farm management strategies suitable for optimum production of vegetables. In our previous study, vegetable growth estimation models were developed considering the whole crop in the each plot as one crop. The previous method had difficulty in obtaining a lot of data, and the reliability of the models was poor. This study reports on the biophysical parameter modeling and validation of the models for white radish and napa cabbage on the basis of UAV-based RGB images by applying new ROIs, which could separate crops in small-bulk shape. Specific objective was to investigate the potential of the UAV-based RGB camera system for effectively quantifying temporal and spatial variability in the growth status of white radish and Chinese cabbage in a field. Otsu threshold-based vegetation fraction and DSM-based plant height were used as two predictor variables for multiple linear regression models to estimate biophysical parameters. Most of the developed models showed a high correlation with R² > 0.7. All validation results showed a high correlation between actual and estimated parameters with R² > 0.79. These meant that the vegetation fraction and plant height were very robust parameters for biophysical parameter models in white radish and napa cabbage. The performances of the biophysical parameter models were significantly reliable to quantify temporal and spatial biophysical parameters.