A deep learning generative model approach for image synthesis of plant leaves.

OBJECTIVES: A well-known drawback to the implementation of Convolutional Neural Networks (CNNs) for image-recognition is the intensive annotation effort for large enough training dataset, that can become prohibitive in several applications. In this study we focus on applications in the agricultural domain and we implement Deep Learning (DL) techniques for the automatic generation of meaningful synthetic images of plant leaves, which can be used as a virtually unlimited dataset to train or validate specialized CNN models or other image-recognition algorithms. METHODS: Following an approach based on DL generative models, we introduce a Leaf-to-Leaf Translation (L2L) algorithm, able to produce collections of novel synthetic images in two steps: first, a residual variational autoencoder architecture is used to generate novel synthetic leaf skeletons geometry, starting from binarized skeletons obtained from real leaf images. Second, a translation via Pix2pix framework based on conditional generator adversarial networks (cGANs) reproduces the color distribution of the leaf surface, by preserving the underneath venation pattern and leaf shape. RESULTS: The L2L algorithm generates synthetic images of leaves with meaningful and realistic appearance, indicating that it can significantly contribute to expand a small dataset of real images. The performance was assessed qualitatively and quantitatively, by employing a DL anomaly detection strategy which quantifies the anomaly degree of synthetic leaves with respect to real samples. Finally, as an illustrative example, the proposed L2L algorithm was used for generating a set of synthetic images of healthy end diseased cucumber leaves aimed at training a CNN model for automatic detection of disease symptoms. CONCLUSIONS: Generative DL approaches have the potential to be a new paradigm to provide low-cost meaningful synthetic samples. Our focus was to dispose of synthetic leaves images for smart agriculture applications but, more in general, they can serve for all computer-aided applications which require the representation of vegetation. The present L2L approach represents a step towards this goal, being able to generate synthetic samples with a relevant qualitative and quantitative resemblance to real leaves.

Assessment of Comfort Variation among Different Types of Driving Agricultural Tractors: Traditional, Satellite-Assisted and Semi-Automatic.

Over the past years, in the agricultural field, geo-localization has been introduced in order to develop specific farming processes, optimize resources, and reduce environmental pollution. Researchers have found alternative driving methods to traditional ones, such as assisted and semi-automatic driving. The aim of this study was to monitor the musculoskeletal efforts necessary to carry out different kinds of driving. The muscular strain was assessed using surface electromyographic devices, the distribution of the pressure exerted by the operator's body on the seat was observed by using two barometric pads applied on the seat back and on the seat, respectively, while the body movements and postures were analyzed through a Microsoft Kinect Camera 3D acquisition system. Results showed a significantly greater muscular activation during manual and assisted driving conditions. The pressure exerted by the operator on the barometric pads was significantly higher in manual and semi-automatic driving modes than in the assisted one. A remarkable increase in the average swinging speed of examined joints was also detected, as well as the distances run by the joints in semi-automatic driving. From our study, assisted driving seems to be the best driving mode both in terms of joint economy and from the efficiency of agricultural processes.

Characterization of the Biogenic Volatile Organic Compounds (BVOCs) and Analysis of the PR1 Molecular Marker in Vitis vinifera L. Inoculated with the Nematode Xiphinema index.

Upon pathogen attack, plants very quickly undergo rather complex physico-chemical changes, such as the production of new chemicals or alterations in membrane and cell wall properties, to reduce disease damages. An underestimated threat is represented by root parasitic nematodes. In Vitis vinifera L., the nematode Xiphinema index is the unique vector of Grapevine fanleaf virus, responsible for fanleaf degeneration, one of the most widespread and economically damaging diseases worldwide. The aim of this study was to investigate changes in the emission of biogenic volatile organic compounds (BVOCs) in grapevines attacked by X. index. BVOCs play a role in plant defensive mechanisms and are synthetized in response to biotic damages. In our study, the BVOC profile was altered by the nematode feeding process. We found a decrease in ß-ocimene and limonene monoterpene emissions, as well as an increase in α-farnesene and α-bergamotene sesquiterpene emissions in nematode-treated plants. Moreover, we evaluated the PR1 gene expression. The transcript level of PR1 gene was higher in the nematode-wounded roots, while in the leaf tissues it showed a lower expression compared to control grapevines.

Evaluation of Borage Extracts As Potential Biostimulant Using a Phenomic, Agronomic, Physiological, and Biochemical Approach.

Biostimulants are substances able to improve water and nutrient use efficiency and counteract stress factors by enhancing primary and secondary metabolism. Premise of the work was to exploit raw extracts from leaves (LE) or flowers (FE) of Borago officinalis L., to enhance yield and quality of Lactuca sativa 'Longifolia,' and to set up a protocol to assess their effects. To this aim, an integrated study on agronomic, physiological and biochemical aspects, including also a phenomic approach, has been adopted. Extracts were diluted to 1 or 10 mL L-1, sprayed onto lettuce plants at the middle of the growing cycle and 1 day before harvest. Control plants were treated with water. Non-destructive analyses were conducted to assess the effect of extracts on biomass with an innovative imaging technique, and on leaf photosynthetic efficiency (chlorophyll a fluorescence and leaf gas exchanges). At harvest, the levels of ethylene, photosynthetic pigments, nitrate, and primary (sucrose and total sugars) and secondary (total phenols and flavonoids) metabolites, including the activity and levels of phenylalanine ammonia lyase (PAL) were assessed. Moreover, a preliminary study of the effects during postharvest was performed. Borage extracts enhanced the primary metabolism by increasing leaf pigments and photosynthetic activity. Plant fresh weight increased upon treatments with 10 mL L-1 doses, as correctly estimated by multi-view angles images. Chlorophyll a fluorescence data showed that FEs were able to increase the number of active reaction centers per cross section; a similar trend was observed for the performance index. Ethylene was three-fold lower in FEs treatments. Nitrate and sugar levels did not change in response to the different treatments. Total flavonoids and phenols, as well as the total protein levels, the in vitro PAL specific activity, and the levels of PAL-like polypeptides were increased by all borage extracts, with particular regard to FEs. FEs also proved efficient in preventing degradation and inducing an increase in photosynthetic pigments during storage. In conclusion, borage extracts, with particular regard to the flower ones, appear to indeed exert biostimulant effects on lettuce; future work will be required to further investigate on their efficacy in different conditions and/or species.

Analysis of electric energy consumption of automatic milking systems in different configurations and operative conditions.

Automatic milking systems (AMS) have been a revolutionary innovation in dairy cow farming. Currently, more than 10,000 dairy cow farms worldwide use AMS to milk their cows. Electric consumption is one of the most relevant and uncontrollable operational cost of AMS, ranging between 35 and 40% of their total annual operational costs. The aim of the present study was to measure and analyze the electric energy consumption of 4 AMS with different configurations: single box, central unit featuring a central vacuum system for 1 cow unit and for 2 cow units. The electrical consumption (daily consumption, daily consumption per cow milked, consumption per milking, and consumption per 100L of milk) of each AMS (milking unit + air compressor) was measured using 2 energy analyzers. The measurement period lasted 24h with a sampling frequency of 0.2Hz. The daily total energy consumption (milking unit + air compressor) ranged between 45.4 and 81.3 kWh; the consumption per cow milked ranged between 0.59 and 0.99 kWh; the consumption per milking ranged between 0.21 and 0.33 kWh; and the consumption per 100L of milk ranged between 1.80 to 2.44 kWh according to the different configurations and operational contexts considered. Results showed that AMS electric consumption was mainly conditioned by farm management rather than machine characteristics/architectures.

Testing of a simplified LED based vis/NIR system for rapid ripeness evaluation of white grape (Vitis vinifera L.) for Franciacorta wine.

The aim of this work was to test a simplified optical prototype for a rapid estimation of the ripening parameters of white grape for Franciacorta wine directly in field. Spectral acquisition based on reflectance at four wavelengths (630, 690, 750 and 850 nm) was proposed. The integration of a simple processing algorithm in the microcontroller software would allow to visualize real time values of spectral reflectance. Non-destructive analyses were carried out on 95 grape bunches for a total of 475 berries. Samplings were performed weekly during the last ripening stages. Optical measurements were carried out both using the simplified system and a portable commercial vis/NIR spectrophotometer, as reference instrument for performance comparison. Chemometric analyses were performed in order to extract the maximum useful information from optical data. Principal component analysis (PCA) was performed for a preliminary evaluation of the data. Correlations between the optical data matrix and ripening parameters (total soluble solids content, SSC; titratable acidity, TA) were carried out using partial least square (PLS) regression for spectra and using multiple linear regression (MLR) for data from the simplified device. Classification analysis were also performed with the aim of discriminate ripe and unripe samples. PCA, MLR and classification analyses show the effectiveness of the simplified system in separating samples among different sampling dates and in discriminating ripe from unripe samples. Finally, simple equations for SSC and TA prediction were calculated.

A Simplified, Light Emitting Diode (LED) Based, Modular System to be Used for the Rapid Evaluation of Fruit and Vegetable Quality: Development and Validation on Dye Solutions.

NIR spectroscopy has proven to be one of the most efficient and ready to transfer tools to monitor product's quality. Portable VIS/NIR instruments are particularly versatile and suitable for field use to monitor the ripening process or quality parameters. The aim of this work is to develop and evaluate a new simplified optoelectronic system for potential measurements on fruit and vegetables directly in the field. The development, characterization and validation of an operative prototype is discussed. LED technology was chosen for the design, and spectral acquisition at four specific wavelengths (630, 690, 750 and 850 nm) was proposed. Nevertheless, attention was given to the modularity and versatility of the system. Indeed, the possibility to change the light sources module with other wavelengths allows one to adapt the use of the same device for different foreseeable applications and objectives, e.g., ripeness evaluation, detection of particular diseases and disorders, chemical and physical property prediction, shelf life analysis, as well as for different natures of products (berry, leaf or liquid). Validation tests on blue dye water solutions have shown the capability of the system of discriminating low levels of reflectance, with a repeatability characterized by a standard deviation proportional to the measured intensity and in general limited to 2%-4%.

Microbial community structure and dynamics in two-stage vs single-stage thermophilic anaerobic digestion of mixed swine slurry and market bio-waste.

The microbial community of a thermophilic two-stage process was monitored during two-months operation and compared to a conventional single-stage process. Qualitative and quantitative microbial dynamics were analysed by Denaturing Gradient Gel Electrophoresis (DGGE) and real-time PCR techniques, respectively. The bacterial community was dominated by heat-shock resistant, spore-forming clostridia in the two-stage process, whereas a more diverse and dynamic community (Firmicutes, Bacteroidetes, Synergistes) was observed in the single-stage process. A significant evolution of bacterial community occurred over time in the acidogenic phase of the two-phase process with the selection of few dominant species associated to stable hydrogen production. The archaeal community, dominated by the acetoclastic Methanosarcinales in both methanogen reactors, showed a significant diversity change in the single-stage process after a period of adaptation to the feeding conditions, compared to a constant stability in the methanogenic reactor of the two-stage process. The more diverse and dynamic bacterial and archaeal community of single-stage process compared to the two-stage process accounted for the best degradation activity, and consequently the best performance, in this reactor. The microbiological perspective proved a useful tool for a better understanding and comparison of anaerobic digestion processes.

Two-stage vs single-stage thermophilic anaerobic digestion: comparison of energy production and biodegradation efficiencies.

Two-stage anaerobic digestion (AD) for integrated biohydrogen and biomethane production from organic materials has been reported to promise higher process efficiency and energy recoveries as compared to traditional one-stage AD. This work presents a comparison between two-stage (reactors R1 and R2) and one-stage (reactor R3) AD systems, fed with identical organic substrates and loading rates, focusing the attention on chemical and microbiological aspects. Contrary to previous experiences, no significant differences in overall energy recovery were found for the two-stage and one-stage AD systems. However, an accumulation in R2 of undegraded intermediate metabolites (volatile fatty acids, ketones, amines, amino acids, and phenols) was observed by GC-MS. These compounds were thought to be both cause and effect of this partial inefficiency of the two-stage system, as confirmed also by the less diverse, and thereby less efficient, population of fermentative bacteria observed (by PCR-DGGE) in R2. The extreme environment of R1 (low pH and high metabolites concentrations) probably acted as selector of metabolic pathways, favoring H(2)-producing bacteria able to degrade such a wide variability of intermediate metabolites while limiting other strains. Therefore, if two-stage AD may potentially lead to higher energy recoveries, further efforts should be directed to ensure process efficiency and stability.

The potential of optical canopy measurement for targeted control of field crop diseases.

There is increasing pressure to reduce the use of pesticides in modern crop production to decrease the environmental impact of current practice and to lower production costs. It is therefore imperative that sprays are only applied when and where needed. Since diseases in fields are frequently patchy, sprays may be applied unnecessarily to disease-free areas. Disease control could be more efficient if disease patches within fields could be identified and spray applied only to the infected areas. Recent developments in optical sensor technology have the potential to enable direct detection of foliar disease under field conditions. This review assesses recent developments in the use of optical methods for detecting foliar disease, evaluates the likely benefits of spatially selective disease control in field crops, and discusses practicalities and limitations of using optical disease detection systems for crop protection in precision pest management.