Search

The leaf area index (LAI) is a biophysical crop parameter of great interest for agronomists and plant breeders. Direct methods for measuring LAI are normally destructive, while indi rect methods are either costly or require long pre- and post-processing times. In this study, a novel deep learning-based (DL) model was developed using RGB nadir-view images taken from a high-throughput plant phenotyping platform for LAI estimation of maize. The study took place in a commercial maize breeding trial during two consecutive grow ing seasons. Ground-truth LAI values were obtained non-destructively using an allometric relationship that was derived to calculate the leaf area of individual leaves from their main leaf dimensions (length and maximum width). Three convolutional neural network (CNN)- based DL model approaches were proposed using RGB images as input. One of the models tested is a classifcation model trained with a set of RGB images tagged with previously measured LAI values (classes). The second model provides LAI estimates from CNN based linear regression and the third one uses a combination of RGB images and numeri cal data as input of the CNN-based model (multi-input model). The results obtained from the three approaches were compared against ground-truth data and LAI estimations from a classic indirect method based on nadir-view image analysis and gap fraction theory. All DL approaches outperformed the classic indirect method. The multi-input_model showed the least error and explained the highest proportion of the observed LAI variance. This work represents a major advance for LAI estimation in maize breeding plots as compared to pre vious methods, in terms of processing time and equipment costs.

Current land surface schemes in weather and climate models make use of the so-called coupled photosynthesis–stomatal conductance (A–gs) models of plant function to determine the surface fluxes that govern the terrestrial energy, water and carbon budgets. Plant physiology is controlled by many environmental factors, and a number of complex feedbacks are involved, but soil moisture control on root water uptake is primary, particularly in sub-tropical to temperate ecosystems. Land surface models represent plant water stress in different ways, but most implement a water stress factor, β, which ranges linearly (more recently also curvilinearly) between β = 1 for unstressed vegetation and β = 0 at the wilting point, expressed in terms of volumetric water content ((Formula presented.)). (Formula presented.) is most commonly used to either limit A or gs, and hence carbon and water fluxes, and a pertinent research question is whether these treatments are in fact interchangeable. Following Egea et al. (Agricultural and Forest Meteorology, 2011, 151 (10), 1,370–1,384) and Verhoef et al. (Agricultural and Forest Meteorology, 2014, 191, 22–32), we have implemented new β treatments, reflecting higher levels of biophysical complexity in a state-of-the-art LSM, Joint UK Land Environment Simulator, by allowing root zone soil moisture to limit plant function non-linearly and via individual routes (carbon assimilation, stomatal conductance, or mesophyll conductance) as well as any (non-linear) combinations thereof. The treatment of β does matter to the prediction of water and carbon fluxes: this study demonstrates that it represents a key structural uncertainty in contemporary LSMs, in terms of predictions of gross primary productivity, energy fluxes and soil moisture evolution, both in terms of climate means and response to a number of European droughts, including the 2003 heat wave. Treatments allowing ß to act on vegetation fluxes via stomatal and mesophyll

Accurate and early estimation of citrus yields is important for both producers and agricultural cooperatives to be competitive and make informed decisions when selling their products. Yield estimation is key for predicting stock volumes, avoiding stock ruptures and planning harvesting operations. Visual yield estimations have tra ditionally been employed, resulting in inaccurate and misleading information. The main goal of this study was to develop an automated image processing methodology to detect, count and estimate the size of citrus fruits on individual trees using deep learning techniques. During 3 consecutive annual campaigns, a total of 20 trees from a commercial citrus grove were monitored using images captured from an unmanned aerial vehicle (UAV). These trees were harvested manually, and fruit sizes were measured. A Faster R-CNN Deep Learning model was trained using a custom dataset to detect oranges in the obtained images. An average standard error (SE) of 6.59 % was obtained between visual counting and the model’s fruit detection. Using the detected fruits, fruit size estimation was also performed. The promising results obtained indicate that this size estimation method can be employed for size discrimination prior to harvest. A model based on Long Short-term Memory (LSTM) was trained for yield estimation per tree and for a total yield estimation. The actual and estimated yields per tree were compared, resulting in an approximate error of SE = 4.53 % and a standard deviation of SD = 0.97 Kg. The actual total yield, the estimated total yield and the total yield estimated by an expert technician were compared. The error in the estimation by the technician was SE = 13.74 %, while the errors in the model were SE = 7.22 % and SD = 4083.58 Kg. These promising results demonstrate the potential of the present technique to provide yield esti mates for citrus fruits or even other types of fruit.

The use of reliable information and data that are rapidly and easily acquired is essential for farm water management and appropriate irrigation strategies. Over the past decade, new advances have been made in irrigation technology, such as platforms that continuously transmit data between irrigation controllers and field sensors, mobile apps, and equipment for variable rate irrigation. In this study, images captured with a thermal imaging camera mounted on an unmanned aerial vehicle (UAV) were used to evaluate the water status of sugar beet plants in a plot with large spatial variability in terms of soil properties. The results were compared with those of soil moisture measurements. No direct relationship was observed between the water status of the soil and that of the crops. However, the fresh root mass and sugar content tended to decrease when higher levels of water stress were detected in the crop using thermal imaging, with coefficients of determination of 0.28 and 0.94 for fresh root mass and sugar content, respectively. Differences were observed be tween different soil types, and therefore different irrigation strategies are needed in highly heterogeneous plots.