Accurate quantification of the spatial variation of canopy size is crucial for vineyard management in the context of Precision Viticulture. Biophysical parameters associated with canopy size, such as ...Leaf Area Index (LAI), can be estimated from Vegetation Indices (VI) such as the Normalized Difference Vegetation Index (NDVI), but in Vertical-Shoot-Positioned (VSP) vineyards, common satellite, or aerial imagery with moderate-resolution capture information at nadir of pixels whose values are a mix of canopy, sunlit soil, and shaded soil fractions and their respective spectral signatures. VI values for each fraction are considerably different. On a VSP vineyard, the illumination direction for each specific row orientation depends on the relative position of sun and earth. Respective proportions of shaded and sunlit soil fractions change as a function of solar elevation and azimuth, but canopy fraction is independent of these variations. The focus of this study is the interaction of illumination direction with canopy orientation, and the corresponding effect on integrated NDVI. The results confirm that factors that intervene in determining the direction of illumination on a VSP will alter the integrated NDVI value. Shading induced considerable changes in the NDVI proportions affecting the final integrated NDVI value. However, the effect of shading decreases as the row orientation approaches the solar path. Therefore, models of biophysical parameters using moderate-resolution imagery should consider corrections for variations caused by factors affecting the angle of illumination to provide more general solutions that may enable canopy data to be obtained from mixed, integrated vine NDVI.
The use of Unmanned Aerial Vehicles (UAVs) in viticulture permits the capture of aerial Red-Green-Blue (RGB) images with an ultra-high spatial resolution. Recent studies have demonstrated that RGB ...images can be used to monitor spatial variability of vine biophysical parameters. However, for estimating these parameters, accurate and automated segmentation methods are required to extract relevant information from RGB images. Manual segmentation of aerial images is a laborious and time-consuming process. Traditional classification methods have shown satisfactory results in the segmentation of RGB images for diverse applications and surfaces, however, in the case of commercial vineyards, it is necessary to consider some particularities inherent to canopy size in the vertical trellis systems (VSP) such as shadow effect and different soil conditions in inter-rows (mixed information of soil and weeds). Therefore, the objective of this study was to compare the performance of four classification methods (K-means, Artificial Neural Networks (ANN), Random Forest (RForest) and Spectral Indices (SI)) to detect canopy in a vineyard trained on VSP. Six flights were carried out from post-flowering to harvest in a commercial vineyard cv. Carménère using a low-cost UAV equipped with a conventional RGB camera. The results show that the ANN and the simple SI method complemented with the Otsu method for thresholding presented the best performance for the detection of the vine canopy with high overall accuracy values for all study days. Spectral indices presented the best performance in the detection of Plant class (Vine canopy) with an overall accuracy of around 0.99. However, considering the performance pixel by pixel, the Spectral indices are not able to discriminate between Soil and Shadow class. The best performance in the classification of three classes (Plant, Soil, and Shadow) of vineyard RGB images, was obtained when the SI values were used as input data in trained methods (ANN and RForest), reaching overall accuracy values around 0.98 with high sensitivity values for the three classes.
Leaf area per unit surface (LAI—leaf area index) is a valuable parameter to assess vine vigour in several applications, including direct mapping of vegetative–reproductive balance (VRB). Normalized ...difference vegetation index (NDVI) has been successfully used to assess the spatial variability of estimated LAI. However, sometimes NDVI is unsuitable due to its lack of sensitivity at high LAI values. Moreover, the presence of hail protection with Grenbiule netting also affects incident light and reflection, and consequently spectral response. This study analyses the effect of protective netting in the LAI–NDVI relationship and, using NDVI as a reference index, compares several indices in terms of accuracy and sensitivity using linear and logarithmic models. Among the indices compared, results show NDVI to be the most accurate, and ratio vegetation index (RVI) to be the most sensitive. The wide dynamic range vegetation index (WDRVI) presented a good balance between accuracy and sensitivity. Soil-adjusted vegetation index 2 (SAVI2) appears to be the best estimator of LAI with linear models. Logarithmic models provided higher determination coefficients, but this has little influence over the normal range of LAI values. A similar NDVI–LAI relationship holds for protected and unprotected canopies in initial vegetation stages, but different functions are preferable once the canopy is fully developed, in particular, if tipping is performed.
•Two new automatic models for estimating reference temperatures are proposed.•Only conventional weather station data are needed for the proposed models.•Heat transfer and empirical models were well ...correlated with the field measurements.•The proposed models can provide accurate reference temperature values.
The use of temperature as an indicator of water stress is gaining attention in agriculture. However, the need for reference temperature measurements in the field (Twet and Tdry) for normalization purposes (calculation of the crop water stress index - CWSI) inhibits the practical implementation of this technique. Therefore, in this study, two new models namely the heat transfer (HT) model and the empirical (EMP) model, are presented and compared with the standard method of physical reference temperature measurements and leaf energy balance calculation as an exploratory case study on grapevines. The HT model is a novel method, based on physical heat transfer principles and uses only input data obtained from a conventional weather station to determine reference temperatures. The EMP model is a simple method based on wet- and dry-bulb temperatures, calculated from ambient temperature and relative humidity measurements. To develop and evaluate the new models, physical measurements of the reference temperatures were taken in a commercial vineyard cv. Cabernet Sauvignon under different levels of water stress at two times during the day on seven days over the growing season. These physical measurements were used to optimize unknown parameters in the new methods by using the particle swarm optimization procedure. Input data for the model was collected by a conventional weather station located nearby the experimental vineyard block. In the validation process, it was found that the HT model can accurately predict the reference temperatures to within 0.5 °C and 1.0 °C for Twet and Tdry, respectively, reacting to environmental conditions as expected. The EMP model, requiring even less meteorological information, can accurately predict the reference temperatures to within 0.8 °C and 1.4 °C for Twet and Tdry, respectively. These proposed methods can provide reference temperatures that do not require physical measurements which can make the use of CWSI more practical and easier to implement for determining plant water stress in vineyards.
Vineyard yield estimation provides the winegrower with insightful information regarding the expected yield, facilitating managerial decisions to achieve maximum quantity and quality and assisting the ...winery with logistics. The use of proximal remote sensing technology and techniques for yield estimation has produced limited success within viticulture. In this study, 2-D RGB and 3-D RGB-D (Kinect sensor) imagery were investigated for yield estimation in a vertical shoot positioned (VSP) vineyard. Three experiments were implemented, including two measurement levels and two canopy treatments. The RGB imagery (bunch- and plant-level) underwent image segmentation before the fruit area was estimated using a calibrated pixel area. RGB-D imagery captured at bunch-level (mesh) and plant-level (point cloud) was reconstructed for fruit volume estimation. The RGB and RGB-D measurements utilised cross-validation to determine fruit mass, which was subsequently used for yield estimation. Experiment one's (laboratory conditions) bunch-level results achieved a high yield estimation agreement with RGB-D imagery (r
= 0.950), which outperformed RGB imagery (r
= 0.889). Both RGB and RGB-D performed similarly in experiment two (bunch-level), while RGB outperformed RGB-D in experiment three (plant-level). The RGB-D sensor (Kinect) is suited to ideal laboratory conditions, while the robust RGB methodology is suitable for both laboratory and in-situ yield estimation.
Research and innovation activities in the area of sensor technology can accelerate the adoption of new and emerging digital tools in the agricultural sector by the implementation of precision farming ...practices such as remote sensing, operations, and real-time monitoring ...
Improving wine composition is a critical factor for the wine industry. Phenolic compounds play an important role in wine composition contributing to its organoleptic characteristics. Although several ...factors can influence the phenolic concentration, plant water status in particular has shown to have a direct impact on the phenolic compounds. It is however complex to quantitate water deficit by plant water status measurements as they depend on the specific site (topography, viticultural management practices and soil characteristics) creating variable values within the vineyard block. This study focused on analysing the effect of natural spatial and temporal variability of plant water status on grape and wine parameters. A field experiment was done in a commercial Cabernet Sauvignon block to monitor the temporal and spatial intra-block variability of plant water status using a grid sample method. Soil analysis and topography were included in the evaluation. Each target vine was assessed for yield, ripeness as well as standard juice parameters. Micro-vinification was done for each target vine and the concentration of anthocyanins and tannins analysed. The results showed that the spatial and temporal variability was evident along the season. Plant water status influenced changes in the concentration of phenolic compounds and grape parameters. The vines in the stressed class were associated with changes in soil texture and topography. These plants presented a moderate increase (6.7%) in sugar content; a significant increase in anthocyanins (22.2%) and tannins (27.5%); and a strong reduction in yield (53.2%) when compared with the non-stressed classes. The results of this study may help to understand and quantify how spatial variability is naturally distributed and its effect on grape and wine parameters.
•Several factors can influence the phenolic concentration in wine; however, water stress has a direct impact.•Plant water status is variable with-in the vineyard block and depend on the specific site conditions.•Temporal and spatial intra-block variability of plant water status was monitored using a grid sample method.•The level of water stress reached by the vines was associated with changes in soil texture, topography, and climatic conditions along the season.•The vines in the stressed class presented a significant increase in sugar content, anthocyanins and tannins and a strong decrease in yield.
The determination of bunch features that are relevant for bunch weight estimation is an important step in automatic vineyard yield estimation using image analysis. The conversion of 2D image features ...into mass can be highly dependent on grapevine cultivar, as the bunch morphology varies greatly. This paper aims to explore the relationships between bunch weight and bunch features obtained from image analysis considering a multicultivar approach. A set of 192 bunches from four cultivars, collected at sites located in Portugal and South Africa, were imaged using a conventional digital RGB camera, followed by image analysis, where several bunch features were extracted, along with physical measurements performed in laboratory conditions. Image data features were explored as predictors of bunch weight, individually and in a multiple stepwise regression analysis, which were then tested on 37% of the data. The results show that the variables bunch area and visible berries are good predictors of bunch weight (R2 ranging from 0.72 to 0.90); however, the simple regression lines fitted between these predictors and the response variable presented significantly different slopes among cultivars, indicating cultivar dependency. The elected multiple regression model used a combination of four variables: bunch area, bunch perimeter, visible berry number, and average berry area. The regression analysis between the actual and estimated bunch weight yielded a R2 = 0.91 on the test set. Our results are an important step towards automatic yield estimation in the vineyard, as they increase the possibility of applying image-based approaches using a generalized model, independent of the cultivar.
Commercial harvest maturity of ‘Hass’ avocado fruit is estimated based on dry matter content (DM). Typically, a few samples representing the entire orchard are destructively analysed using ...time-consuming procedures such as oven or freeze drying the fruit's mesocarp. However, the maturity parameter of avocado, that is known to have a direct link to nutritional quality, is oil content (OC). This study was conducted to develop models for indexing maturity of on-tree avocado using a portable visible to near-infrared spectrometer. Rapid non-destructive models for assessing OC, DM and moisture content (MC) of avocado fruit were successfully developed using The Unscrambler® X chemometric software. Models robustness was assessed in an independent test set. There were non-significant differences (p > 0.05) between destructive and non-destructively assessed OC in terms of means (42.45 and 41.91%), standard deviations (4.79 and 4.87%) and coefficients of variation (11.34 and 11.62%) from the independent test set. The predictability of OC was associated with its high extractability caused by drying samples at high (75 °C) temperatures. The heat-drying technique can be used by other researchers to increase extractability and hence, the predictability of avocado OC during calibrations of alike non-destructive models. Commercial application of the developed models can improve maturity indexing since OC, DM and MC can be easily assessed without harvesting of sample fruit.
Display omitted
•Vis–NIRS successfully quantified oil content (OC) of on-tree avocado fruit.•Oven-drying samples at 70 °C improved the predictability of the fruit OC.•Application of spectrometry can increase sample throughput for maturity indexing.•Applying portable spectrometers can increase the number of reference indices.•Assessing on-tree samples was linked with avoiding unnecessary waste of fruit.