Over the last few years, several Convolutional Neural Networks for object detection have been proposed, characterised by different accuracy and speed. In viticulture, yield estimation and prediction is used for efficient crop management, taking advantage of precision viticulture techniques. Convolutional Neural Networks for object detection represent an alternative methodology for grape yield estimation, which usually relies on manual harvesting of sample plants. In this paper, six versions of the You Only Look Once (YOLO) object detection algorithm (YOLOv3, YOLOv3-tiny, YOLOv4, YOLOv4-tiny, YOLOv5x, and YOLOv5s) were evaluated for real-time bunch detection and counting in grapes. White grape varieties were chosen for this study, as the identification of white berries on a leaf background is trickier than red berries. YOLO models were trained using a heterogeneous dataset populated by images retrieved from open datasets and acquired on the field in several illumination conditions, background, and growth stages. Results have shown that YOLOv5x and YOLOv4 achieved an F1-score of 0.76 and 0.77, respectively, with a detection speed of 31 and 32 FPS. Differently, YOLO5s and YOLOv4-tiny achieved an F1-score of 0.76 and 0.69, respectively, with a detection speed of 61 and 196 FPS. The final YOLOv5x model for bunch number, obtained considering bunch occlusion, was able to estimate the number of bunches per plant with an average error of 13.3% per vine. The best combination of accuracy and speed was achieved by YOLOv4-tiny, which should be considered for real-time grape yield estimation, while YOLOv3 was affected by a False Positive–False Negative compensation, which decreased the RMSE.

Monitoring and prediction of within-field crop variability can support farmers to make the right decisions in different situations. The current advances in remote sensing and the availability of high resolution, high frequency, and free Sentinel-2 images improve the implementation of Precision Agriculture (PA) for a wider range of farmers. This study investigated the possibility of using vegetation indices (VIs) derived from Sentinel-2 images and machine learning techniques to assess corn (Zea mays) grain yield spatial variability within the field scale. A 22-ha study field in North Italy was monitored between 2016 and 2018; corn yield was measured and recorded by a grain yield monitor mounted on the harvester machine recording more than 20,000 georeferenced yield observation points from the study field for each season. VIs from a total of 34 Sentinel-2 images at different crop ages were analyzed for correlation with the measured yield observations. Multiple regression and two different machine learning approaches were also tested to model corn grain yield. The three main results were the following: (i) the Green Normalized Difference Vegetation Index (GNDVI) provided the highest R² value of 0.48 for monitoring within-field variability of corn grain yield; (ii) the most suitable period for corn yield monitoring was a crop age between 105 and 135 days from the planting date (R4–R6); (iii) Random Forests was the most accurate machine learning approach for predicting within-field variability of corn yield, with an R² value of almost 0.6 over an independent validation set of half of the total observations. Based on the results, within-field variability of corn yield for previous seasons could be investigated from archived Sentinel-2 data with GNDVI at crop stage (R4–R6).

Aim: The recent availability of Sentinel-2 satellites has led to an increasing interest in their use in viticulture. The aim of this short communication is to determine performance and limitation of a Sentinel-2 vegetation index in precision viticulture applications, in terms of correlation and variability assessment, compared to the same vegetation index derived from an unmanned aerial vehicle (UAV). Normalised difference vegetation index (NDVI) was used as reference vegetation index.Methods and Results: UAV and Sentinel-2 vegetation indices were acquired for 30 vineyard blocks located in the south of France without inter-row grass. From the UAV imagery, the vegetation index was calculated using both a mixed pixels approach (both vine and inter-row) and from pure vine-only pixels. In addition, the vine projected area data were extracted using a support vector machine algorithm for vineyard segmentation. The vegetation index was obtained from Sentinel-2 imagery obtained at approximately the same time as the UAV imagery. The Sentinel-2 images used a mixed pixel approach as pixel size is greater than the row width. The correlation between these three layers and the Sentinel-2 derived vegetation indices were calculated, considering spatial autocorrelation correction for the significance test. The Gini coefficient was used to estimate variability detected by each sensor at the within-field scale. The effects of block border and dimension on correlations were estimated.Conclusions: The comparison between Sentinel-2 and UAV vegetation index showed an increase in correlation when border pixels were removed. Block dimensions did not affect the significance of correlation unless blocks were < 0.5 ha. Below this threshold, the correlation was non-significant in most cases. Sentinel-2 acquired data were strongly correlated with UAV-acquired data at both the field (R2 = 0.87) and sub-field scale (R2 = 0.84). In terms of variability detected, Sentinel-2 proved to be able to detect the same amount of variability as the UAV mixed pixel vegetation index.Significance and impact of the study: This study showed at which field conditions the Sentinel-2 vegetation index can be used instead of UAV-acquired images when high spatial resolution (vine-specific) management is not needed and the vineyard is characterised by no inter-row grass. This type of information may help growers to choose the most appropriate information sources to detect variability according to their vineyard characteristics.

Mapping crop within-field yield variability provide an essential piece of information for precision agriculture applications. Leaf Area Index (LAI) is an important parameter that describes maize growth, vegetation structure, light absorption and subsequently maize biomass and grain yield (GY). The main goal for this study was to estimate maize biomass and GY through LAI retrieved from hyperspectral aerial images using a PROSAIL model inversion and compare its performance with biomass and GY estimations through simple vegetation index approaches. This study was conducted in two separate maize fields of 12 and 20 ha located in north-west Mexico. Both fields were cultivated with the same hybrid. One field was irrigated by a linear pivot and the other by a furrow irrigation system. Ground LAI data were collected at different crop growth stages followed by maize biomass and GY at the harvesting time. Through a weekly/biweekly airborne flight campaign, a total of 19 mosaics were acquired between both fields with a micro-hyperspectral Vis-NIR imaging sensor ranging from 400 to 850 nanometres (nm) at different crop growth stages. The PROSAIL model was calibrated and validated for retrieving maize LAI by simulating maize canopy spectral reflectance based on crop-specific parameters. The model was used to retrieve LAI from both fields and to subsequently estimate maize biomass and GY. Additionally, different vegetation indices were calculated from the aerial images to also estimate maize yield and compare the indices with PROSAIL based estimations. The PROSAIL validation to retrieve LAI from hyperspectral imagery showed a R² value of 0.5 against ground LAI with RMSE of 0.8 m²/m². Maize biomass and GY estimation based on NDRE showed the highest accuracies, followed by retrieved LAI, GNDVI and NDVI with R² value of 0.81, 0.73, 0.73 and 0.65 for biomass, and 0.83, 0.69, 0.73 and 0.62 for GY estimation, respectively. Furthermore, the late vegetative growth stage at V16 was found to be the best stage for maize yield prediction for all studied indices.

Defining the most profitable remote sensing platforms is a difficult decision-making process, as it requires agronomic and economic considerations. In this paper, the price and profitability of three levels of remote sensing platforms were evaluated to define a decision-making process. Prices of satellite, plane and UAV-acquired vegetation indices were collected in Italy during 2020 and compared to the economic benefits resulting from variable rate nitrogen application, according to a bibliographic meta-analysis performed on grains. The quality comparison of these three technologies was performed considering the error propagation in the NDVI formula. The errors of the single bands were used to assess the optical properties of the sensors. Results showed that medium-resolution satellite data with good optical properties could be profitably used for variable rate nitrogen applications starting from 2.5 hectares, in case of medium resolution with good optical properties. High-resolution satellites with lower optical quality were profitable starting from 13.2 hectares, while very high-resolution satellites with good optical properties could be profitably used starting from 76.8 hectares. Plane-acquired images, which have good optical properties, were profitable starting from 66.4 hectares. Additionally, a reference model for satellite image price is proposed.