Search

Forcing tulips is a process that mimics the conditions of winter and early spring in a controlled environment to make the tulips bloom months earlier than they would if grown in the open field. In this process, it is essential that the relative humidity in the growing environment is kept sufficiently low. Otherwise, there is an increased risk of the tulips losing solidity and eventually toppling over. Since maintaining this lower humidity demands a lot of energy, it is most cost-efficient to maintain the highest possible relative humidity while still maintaining a rigid flower stem. In order to find the optimal balance between stem rigidity and energy consumption, a monitoring technique is developed that can recognize early signs of the type of high-humidity exposure that eventually leads to toppling. The monitoring technique is based on spectral imaging and a linear classifier that predicts whether the tulips have been forced in an adequately ventilated environment or a poorly ventilated, high-RH enclosure. The model's classification accuracy was found to reach 100% 6-10 days before toppling occurs.

Soil composition and herbivory are two environmental factors that can affect plant traits including flower traits, thus potentially affecting plant–pollinator interactions. Importantly, soil composition and herbivory may interact in these effects, with consequences for plant fitness. We assessed the main effects of aboveground insect herbivory and soil amendment with exuviae of three different insect species on visual and olfactory traits of Brassica nigra plants, including interactive effects. We combined various methodological approaches including gas chromatography/mass spectrometry, spectroscopy and machine learning to evaluate changes in flower morphology, colour and the emission of volatile organic compounds (VOCs). Soil amended with insect exuviae increased the total number of flowers per plant and VOC emission, whereas herbivory reduced petal area and VOC emission. Soil amendment and herbivory interacted in their effect on the floral reflectance spectrum of the base part of petals and the emission of 10 VOCs. These findings demonstrate the effects of insect exuviae as soil amendment on plant traits involved in reproduction, with a potential for enhanced reproductive success by increasing the strength of signals attracting pollinators and by mitigating the negative effects of herbivory.

A novel case of developing a portable spectral imaging device for kiwifruit analysis is presented. Furthermore, a new complementary spectral image processing strategy combining deep learning and advanced chemometric is proposed for processing the spectral images. The deep learning was used for detection and localisation of harvested fruit in the spectral image while the chemometric modelling was used to predict multiple fruit quality related properties i.e., dry matter and soluble solids content. The developed models were independently validated on fruit harvested from a different orchard as well as on a different variety. The one touch spectral imaging presented in this paper can allow widespread usage of spectral imaging for fresh fruit analysis, particularly benefitting non-experts in spectral imaging and chemometrics to routinely use the spectral imaging for fresh fruit analysis.

Purpose of Review A short introduction to the spectral imaging (SI) of plants along with a comprehensive overview of the recent research works related to disease detection in plants using autonomous phenotyping platforms is provided. Key benefits and challenges of SI for plant disease detection on robotic platforms are highlighted. Recent Findings SI is becoming a potential tool for autonomous platforms for non-destructive plant assessment. This is because it can provide information on the plant pigments such as chlorophylls, anthocyanins and carotenoids and supports quantification of biochemical parameters such as sugars, proteins, different nutrients, water and fat content. A plant suffering from diseases will exhibit different physicochemical parameters compared with a healthy plant, allowing the SI to capture those differences as a function of reflected or absorbed light. Summary Potential of SI to non-destructively capture physicochemical parameters in plants makes it a key technique to support disease detection on autonomous platforms. SI can be broadly used for crop disease detection by quantification of physicochemical changes in the plants.

Spectral imaging (SI) in analytical chemistry is widely used for the assessment of spatially distributed physicochemical properties of samples. Although massive development in instrument and chemometrics modelling has taken place in the recent years, the main challenge with SI is that available sensors require extensive system integration and calibration modelling before their use for routine analysis. Further, the models developed during one experiment are rarely useful once the system is reintegrated for a new experiment. To avoid system reintegration and reuse calibrated models, this study presents an intelligent All-In-One SI (ASI) laboratory system allowing standardised automated data acquisition and real-time spectral model deployment. The ASI system supplies a controlled standardised illumination environment, an in-built computing system, embedded software for automated image acquisition, and model deployment to predict the spatial distribution of sample properties in real-time. To show the capability of the ASI framework, exemplary cases of fruit property prediction in different fruits are presented. Furthermore, ASI is also benchmarked in performance against the current commercially available portable as well as high-end laboratory spectrometers.

Potato blackleg is a tuber-borne bacterial disease caused by species within the genera Dickeya and Pectobacterium that can cause decay of plant tissue and wilting through the action of cell wall degrading enzymes released by the pathogen. In case of serious infections, tubers may rot before emergence. Management is largely based on the use of pathogen-free seed potato tubers. For this, fields are visually monitored both for certification and also to take out diseased plants to avoid spread to neighboring plants. Imaging potentially offers a quick and non-destructive way to inspect the health of potato plants in a field. Early detection of blackleg diseased plants with modern vision techniques can significantly reduce costs. In this paper, we studied the use of deep learning for detecting blackleg diseased potato plants. Two deep convolutional neural networks were trained on RGB images with healthy and diseased plants. One of these networks (ResNet18) was experimentally found to produce a precision of 95 % and recall of 91 % for the disease class. These results show that convolutional neural networks can be used to detect blackleg diseased potato plants.

Plant phenotypic plasticity in response to antagonists can affect other community members such as mutualists, conferring potential ecological costs associated with inducible plant defence. For flowering plants, induction of defences to deal with herbivores can lead to disruption of plant–pollinator interactions. Current knowledge on the full extent of herbivore‐induced changes in flower traits is limited, and we know little about specificity of induction of flower traits and specificity of effect on flower visitors. We exposed flowering Brassica nigra plants to six insect herbivore species and recorded changes in flower traits (flower abundance, morphology, colour, volatile emission, nectar quantity, and pollen quantity and size) and the behaviour of two pollinating insects. Our results show that herbivory can affect multiple flower traits and pollinator behaviour. Most plastic floral traits were flower morphology, colour, the composition of the volatile blend, and nectar production. Herbivore‐induced changes in flower traits resulted in positive, negative, or neutral effects on pollinator behaviour. Effects on flower traits and pollinator behaviour were herbivore species‐specific. Flowers show extensive plasticity in response to antagonist herbivores, with contrasting effects on mutualist pollinators. Antagonists can potentially act as agents of selection on flower traits and plant reproduction via plant‐mediated interactions with mutualists., We show that plant interactions with pollinators may be easily disrupted or enhanced by herbivores due to extensive floral plasticity. Herbivore‐species‐specific changes in flower morphology, colour, volatile blend, and nectar and pollen production had contrasting effects on pollinator behaviour.

This study aims to utilize non-destructive sensing based on Vis-NIR spectroscopy and acoustic to predict firmness of avocado fruit. The study has three aims, the first aim was to find the best reference firmness measurement technique for calibrating Vis-NIR spectroscopy data related to avocado ripening i.e., acoustic firmness (AF), limited compression (LC) and penetrometer max force (Fmax). The second aim was to study the generalizability of Vis-NIR models with respect to the dehydration level of avocado fruits. Dehydration of outer skin during storage is common and may cause model failure as the Vis-NIR signal is dominated by signal corresponding to high moisture in fresh fruit. The third aim was to fuse the Vis-NIR spectroscopy and acoustic information to improve the prediction of the LC and Fmax, otherwise unattainable with a single technique. The results showed that the best models for firmness prediction were obtained with LC as the reference. The avocado skin dehydration negatively affected the performance of Vis-NIR models to predict firmness. Further, a fusion of Vis-NIR spectroscopy and acoustic information improved prediction (reduced error by 21%) of firmness in avocado. Assessing avocado firmness in a multi-sensor framework can allow to precisely access the ripeness stage of avocados.

Tomato is an important food product for which the development of non-destructive quality assessment methods is of great interest. Using visible and near-infrared (NIR) spectroscopy, the sugar content, acidity and even taste can be estimated through the use of chemometric methods (e.g., partial least squares regression). In the case of reflection spectra, which are the common modality for imaging spectroscopy, the question arises regarding how much of the interior of the tomato contributes to the measured spectra. An experiment was performed with tomatoes of four different types: beef tomato, classic round tomato, cocktail tomato, and snack tomato. The tomatoes were sliced at different thicknesses and imaged on a 98% reflective white background and a 4% reflective black background. Spectral images were acquired with VNIR (400–1000 nm) and NIR (900–1700 nm) imaging spectrographs. The difference between the spectra with a white and black background was used to determine the relationship between the wavelength and the light penetration depth. The results show that at wavelengths between 600 and 1100 nm, light penetrates the tomatoes up to a distance of 20 mm. The relation more or less follows the law of Lambert–Beer. This relation was the same for all four types of tomatoes. These results help the interpretation of chemometric models based on reflection (imaging) spectroscopy.

International audience; Close-range spectral imaging (SI) of agricultural plants is widely performed for digital plant phenotyping. A key task in digital plant phenotyping is the non-destructive and rapid identification of drought stress in plants so as to allow plant breeders to select potential genotypes for breeding drought-resistant plant varieties. Visible and near-infrared SI is a key sensing technique that allows the capture of physicochemical changes occurring in the plant under drought stress. The main challenges are in processing the massive spectral images to extract information relevant for plant breeders to support genotype selection. Hence, this study presents a generic data processing workflow for analysing SI data generated in real-world digital phenotyping experiments to extract meaningful information for decision making by plant breeders. The workflow is a combination of chemometric approaches and deep learning. The usefulness of the proposed workflow is demonstrated on a real-life experiment related to drought stress detection and quantification in Arabidopsis thaliana plants grown in a semi-controlled environment. The results show that the proposed approach is able to detect the presence of drought just 3 days after its induction compared to the well-watered plants. Furthermore, the unsupervised clustering approach provides detailed time-series images where the drought-related changes in plants can be followed visually along the time course. The developed approach facilitates digital phenotyping and can thus accelerate breeding of drought-tolerant plant varieties.

Accurately detecting and counting fruits during plant growth using imaging and computer vision is of importance not only from the point of view of reducing labor intensive manual measurements of phenotypic information, but also because it is a critical step toward automating processes such as harvesting. Deep learning based methods have emerged as the state-of-the-art techniques in many problems in image segmentation and classification, and have a lot of promise in challenging domains such as agriculture, where they can deal with the large variability in data better than classical computer vision methods. This paper reports results on the detection of tomatoes in images taken in a greenhouse, using the MaskRCNN algorithm, which detects objects and also the pixels corresponding to each object. Our experimental results on the detection of tomatoes from images taken in greenhouses using a RealSense camera are comparable to or better than the metrics reported by earlier work, even though those were obtained in laboratory conditions or using higher resolution images. Our results also show that MaskRCNN can implicitly learn object depth, which is necessary for background elimination.

In greenhouse horticulture, efficiency of climate control and plant protection can be improved by having an accurate impression of plant status, such as photosynthesis or chemical composition. Recent advances in remote sensing technologies have brought about a range of innovations in precision agriculture, with the potential for adaptation to greenhouses. Simple, traditionally used indices employ only one or two spectral bands, in which the contributions of various pigments and leaf or canopy structure can highly overlap. Consequently, such indices may be insufficient for applications. State-of-the-art models have been developed that can better interpret hyper- and multispectral leaf and canopy imagery by employing the biochemical and radiative transfer properties of vegetation. An example is the soil-canopy observation of photosynthesis and energy balance (SCOPE) model, which was developed specifically for crop canopies. Here we present one of the pillars of SCOPE, the leaf radiative transfer (RT) model Fluspect. Fluspect simulates leaf chlorophyll fluorescence, reflectance and transmittance spectra. The model can be inverted to obtain estimates of leaf chlorophylls, carotenoids, anthocyanins, xanthophyll epoxidation, water and dry matter content. Moreover, it can be linked to a model for leaf photosynthesis and when inverted, provide a method to estimate photosynthesis directly from leaf spectral information. We test the model against a tomato data set, with measured hyperspectral images, chlorophyll, sugar, acid, starch, dry matter content and nutrients. The first study of the data set, using partial least square regression, showed that hyperspectral images have a high correlation with important fruit and leaf compounds. We compared these results to Fluspect retrievals and conventional vegetation indices. In the paper, we discuss the potential added value of using RT models in greenhouse horticulture.

Hyperspectral imaging is currently a very well-known and much used technology for measuring features in different fields, such as chemistry, geology, medicine, food and agriculture, either spaceborne (satellites), airborne (drones) or at close proximity (e.g. field scanning, industrial sorting lines or microscopy). Its background is two-fold, and it can be considered as a special case of spectroscopy (“imaging spectroscopy”) or a special case of imaging (“spectral imaging”). Current practice is to use adjectives such as multi and hyper added to “spectral imaging” in order to characterise the number of wavelength bands. In this paper we propose the community to use scientifically sound terminology, like “imaging spectroscopy” or “spectral imaging”, without using ambiguous adjectives. Further, we encourage the community to define and agree upon clear adjectives to describe the number of variables in the naming of our imaging technique.

Plant disease detection represents a tremendous challenge for research and practical applications. Visual assessment by human raters is time-consuming, expensive, and error prone. Disease rating and plant protection need new and innovative techniques to address forthcoming challenges and trends in agricultural production that require more precision than ever before. Within this context, hyperspectral sensors and imaging techniques—intrinsically tied to efficient data analysis approaches—have shown an enormous potential to provide new insights into plant-pathogen interactions and for the detection of plant diseases. This article provides an overview of hyperspectral sensors and imaging technologies for assessing compatible and incompatible plant-pathogen interactions. Within the progress of digital technologies, the vision, which is increasingly discussed in the society and industry, includes smart and intuitive solutions for assessing plant features in plant phenotyping or for making decisions on plant protection measures in the context of precision agriculture.

In this paper we propose an image analysis method for detecting and counting tomato flowers from images taken in a greenhouse. Detecting and locating flowers is useful information for tomato growers and breeders, for phenotyping, yield prediction, and for automating procedures such as pollination and spraying. Since the tomato flowers are yellow, we first apply a set of grayscale transformations in which yellow regions stand out, and then threshold and combine them by a logical binary AND operation. Using more than one transform reduces the possibility of spurious detections due to non-flower regions of the image appearing yellow due to illumination conditions. Connected regions larger than a certain threshold are selected as instances belonging to the class flower. Experimental results over images acquired in a greenhouse using a Realsense camera show that this approach could detect flowers with a recall of 0.79 and precision of 0.77, which are comparable to the values reported in literature with higher resolution cameras closer to the flowers being imaged.

De publiek private samenwerking (PPS) ‘Op naar precisielandbouw 2.0’ (PL2.0) is een R&D project van de topsector AgriFood. Het project is gestart in 2015 met doorlooptijd van 4 jaar. Voor u ligt het eindrapport. In deze PPS werkten ruim 20 private bedrijven en organisaties, publieke kennisinstellingen en overheden samen aan strategische onderwerpen binnen precisielandbouw. Het project omvatte 13 deelprojecten verdeeld over vijf specifieke R&D thema’s, te weten slim satellietbeeldengebruik, sensorontwikkeling (ziektedetectie), slimme integratie van technologieën in toepassingen, perceelkarakteristieken voor schatten van opbrengstpotentie en ondersteunende ICT, en een generiek thema communicatie en kennisverspreiding.Met betrekking tot het thema satellietbeeldengebruik is uitgezocht hoe optische satellietbeelden in combinatie met radarbeelden of beelden verkregen via drone-camera’s beter gebruikt kunnen worden om de variatie en status van de bovengrondse hoeveelheid biomassa van gewassen in kaart te brengen en opbrengsten te voorspellen. Op het gebied van ziektedetectie is door middel van sensor fusion en artificial intelligence de detectie van virus- en bacterieziekten in aardappelplanten verbeterd. En werd een prototype sensorsysteem voor veldonderzoek ontwikkeld. Door slimme integratie van data, adviesmodellen en mechanisatie zijn er enkele variabel-doseertoepassingen ontwikkeld en gevalideerd. Het gaat hier om variabel doseren van Stikstof en herbiciden binnen teelten d.m.v. taakkaarten. In het verlengde hiervan is ook een ontwerp geleverd en als prototype gevalideerd voor een innovatieve beddenspuit in bloembollenteelt. Op grond van perceelkarakteristieken en ondersteunende ICT zijn inzichten en tools voor het inschatten van opbrengst(potentie) geleverd en wordt een doorkijk gegeven naar software voor verbeterde rijpadenplanning en perceelinformatie. De inzet op communicatie en kennisdeling heeft ca. 100 publicaties en presentaties in 4 jaar tijd opgeleverd. Voor meer details over resultaten wordt naar de rapportage met samenvatting per deelproject verwezen in de hoofdstukken 2 tot en met 7.Het grote succes van PL2.0 ligt vooral bij ruime aandacht voor integratie van componenten van precisielandbouwtoepassingen en de doorstroming daarvan naar de praktijk en onderwijs.Geconcludeerd mag worden dat PL2.0 een bijdrage leverde aan gewasmonitoringtoepassingen en diverse variabel-doseertoepassingen (variable rate applications, VRA). Die VRA-toepassingen zien we nu op de agenda in het in 2018 gestarte precisielandbouw-adoptie project ‘Nationale Proeftuin Precisielandbouw’ (NPPL). Meerdere bedrijven passen taakkaarten variabel doseren op een resolutie van 30-50 m2 op praktijkschaal toe en besparen zo’n 20 -30% op gewasbeschermingsmiddelen met behoud van goede werking. De basis hiervoor is een bodem- of gewaskaart die de relevante variatie binnen de bodem of gewas in kaart brengt. Ook zijn er via PL2.0 mooie resultaten met optimalisatie van plantdichtheid en vermindering van meststoffengebruik via deze kaarten. Doorstroming van kennis naar het groene onderwijs werd gerealiseerd via PL2.0 en een versterkend WURKS-traject. Negen lesmodules over gebruik software en inzet taakkaarten in precisielandbouw werden opgeleverd. Precisielandbouw is geen doel op zich, maar een manier om de duurzaamheid van landbouw te vergroten. Met PL2.0 toepassingen kan meer met minder en beter geproduceerd worden. De trend van precisielandbouw c.q. data-gedreven landbouw of smart farming, zal zich alleen maar doorzetten. Er zal gewerkt gaan worden met meer en hoog-resolutie data, complexere adviesmodellen en meer robotisering. Daarmee zullen de doelen van kringlooplandbouw beter en sneller gerealiseerd kunnen worden.

Tulip crop production in the Netherlands suffers from severe economic losses caused by virus diseases such as the Tulip Breaking Virus (TBV). Infected plants which can spread the disease by aphids must be removed from the field as soon as possible. As the availability of human experts for visual inspection in the field is limited, there is an urgent need for a rapid, automated and objective method of screening. From 2009-2012, we developed an automatic machine-vision-based system, using classical machine-learning algorithms. In 2012, the experiment conducted a tulip field planted at production density of 100 and 125 plants per square meter, resulting in images with overlapping plants. Experiments based on multispectral images resulted in scores that approached results obtained by experienced crop experts. The method, however, needed to be tuned specifically for each of the data trails, and a NIR band was needed for background segmentation. Recent developments in artificial intelligence and specifically in the area of convolutional neural networks, allow the development of more generic solutions for the detection of TBV. In this study, a Faster R-CNN network is applied on part of the data from the 2012 experiment. The outcomes show that the results are almost the same compared to the previous method using only RGB data.

Growers would like to know the status of their crop to determine climate strategy and crop management practices. Chemical composition of the crop can now only be determined by sampling leaves or fruits, send this to a laboratory and wait for the analysis. In this project, we aimed to use hyperspectral imaging to determine the contents of sugars and starch, dry matter percentage, chlorophyll and nutrient composition in the crop. The results are promising. Hyperspectral cameras are very well capable to estimate the concentrations of sugars in leaves and fruits, chlorophyll content, dry matter percentage and specific leaf area. This allows the growers to adjust their climate settings or crop management based on these hyperspectral images. Voor telers is het van belang te weten hoeveel reserves hun gewas heeft. Dat kan nu alleen gemeten worden door bladeren of vruchten in te sturen naar een laboratorium en te wachten op de analyse. In dit project is gekeken of het mogelijk is met hyperspectraal camera’s direct gehaltes aan suikers en zetmeel, droge stof, chlorofyl en nutriënten in het gewas te meten. De resultaten zijn veelbelovend: met hyperspectraal camera’s zijn gehaltes aan suikers in bladeren en vruchten, chlorofyl en droge stof percentage en de bladdikte goed te bepalen. Dit maakt het mogelijk om op basis van deze camerabeelden klimaat te sturen of teeltstrategie aan te passen.

Nondestructive methods are of utmost importance for honey characterization. This study investigates the potential application of VIS-NIR hyperspectral imaging for detection of honey flower origin using machine learning techniques. Hyperspectral images of 52 honey samples were taken in transmittance mode in the visible/near infrared (VIS-NIR) range (400–1000 nm). Three different machine learning algorithms were implemented to predict honey floral origin using honey spectral images. These methods, included radial basis function (RBF) network, support vector machine (SVM), and random forest (RF). Principal component analysis (PCA) was also exploited for dimensionality reduction. According to the obtained results, the best classifier (RBF) achieved a precision of 94% in a fivefold cross validation experiment using only the first two PCs. Mapping of the classifier results to the test set images showed 90% accuracy for honey images. Three types of honey including buckwheat, rapeseed and heather were classified with 100% accuracy. The proposed approach has great potential for honey floral origin detection. As some other honey properties can also be predicted using image features, in addition to floral origin detection, this method may be applied to predict other honey characteristics.

In this paper we present a novel method for automated detection of Mycosphaerella melonis infected cucumber fruits. The two-step method consists of machine learning approach using: shape based features extracted from cucumber color images and light transmission spectra based features. The automated detection rate was compared to the manual detection rate of the human workers. Our automated method reached the 95% detection accuracy, which is comparable to the manual detection accuracy of 96%.

Introduction Metabolomics has become a valuable tool in many research areas. However, generating metabolomics-based biochemical profiles without any related bioactivity is only of indirect value in understanding a biological process. Therefore, metabolomics research could greatly benefit from tools that directly determine the bioactivity of the detected compounds. Objective We aimed to combine LC–MS metabolomics with a cell based receptor assay. This combination could increase the understanding of biological processes and may provide novel opportunities for functional metabolomics. Methods We developed a flow through biosensor with human cells expressing both the TRPV1, a calcium ion channel which responds to capsaicin, and the fluorescent intracellular calcium ion reporter, YC3.6. We have analysed three contrasting Capsicum varieties. Two were selected with contrasting degrees of spiciness for characterization by HPLC coupled to high mass resolution MS. Subsequently, the biosensor was then used to link individual pepper compounds with TRPV1 activity. Results Among the compounds in the crude pepper fruit extracts, we confirmed capsaicin and also identified both nordihydrocapsaicin and dihydrocapsaicin as true agonists of the TRPV1 receptor. Furthermore, the biosensor was able to detect receptor activity in extracts of both Capsicum fruits as well as a commercial product. Sensitivity of the biosensor to this commercial product was similar to the sensory threshold of a human sensory panel. Conclusion Our results demonstrate that the TRPV1 biosensor is suitable for detecting bioactive metabolites. Novel opportunities may lie in the development of a continuous functional assay, where the biosensor is directly coupled to the LC–MS. Electronic supplementary material The online version of this article (doi:10.1007/s11306-016-1057-y) contains supplementary material, which is available to authorized users.

This study investigates the application of hyperspectral imaging system and data mining based classifiers for honey adulteration detection. Hyperspectral images from pure and adulterated samples were captured in using a VIS-NIR hyperspectral camera (400 – 1000 nm). After preprocessing the images, five different data mining based techniques, including artificial neural network (ANN), support vector machine (SVM), Linear discriminant analysis (LDA), Fisher and Parzen classifiers were applied for supervised image classification. Classifier test results show the highest classification accuracy of 95% for ANN classifier. Other classifiers including SVM with radial basis kernel function (92%), LDA (90%), Fisher (89 %), and Parzen with 84% correct classification rate also showed acceptable results. This research shows the capability of hyperspectral imaging for honey authentication.

Broad-leaved dock is a common and troublesome grassland weed with a wide geographic distribution. In conventional farming the weed is normally controlled by using a selective herbicide, but in organic farming manual removal is the best option to control this weed. The objective of our work was to develop a robot that can navigate a pasture, detect broad-leaved dock, and remove any weeds found. A prototype robot was constructed that navigates by following a predefined path using centimeter-precision global positioning system (GPS). Broad-leaved dock is detected using a camera and image processing. Once detected, weeds are destroyed by a cutting device. Tests of aspects of the system showed that path following accuracy is adequate but could be improved through tuning of the controller or adoption of a dynamic vehicle model, that the success rate of weed detection is highest when the grass is short and when the broad-leaved dock plants are in rosette form, and that 75% of weeds removed did not grow back. An on-farm field test of the complete system resulted in detection of 124 weeds of 134 encountered (93%), while a weed removal action was performed eight times without a weed being present. Effective weed control is considered to be achieved when the center of the weeder is positioned within 0.1 m of the taproot of the weed—this occurred in 73% of the cases. We conclude that the robot is an effective instrument to detect and control broad-leaved dock under the conditions encountered on a commercial farm. © 2010 Wiley Periodicals, Inc. © 2011 Wiley Periodicals, Inc.

Stress in plants can be measured using chlorophyll fluorescence imaging. The development of patterns in time can give an indication of the type of stress. Since leaves grow and show leaf movements, there is no pixel to pixel correspondence in time laps imaging data. In this article, Penalized Likelihood Warping and Robust Point Matching methods for recovering the pixel to pixel correspondence of a leaf within a time series are studied. It is shown that Robust Point Matching method is more suitable for our application than Penalized Likelihood Warping. Furthermore, Robust Point Matching method is much faster than Penalized Likelihood Warping. After warping an image sequence of a cabbage leaf infected with the bacteria Xanthomonas campestris pv. campestris, it was possible to identify infected spots 30 h after infection, where in unwarped images differences just can be seen 60 h after infection. Time series of the warped image data can be used to study and measure stress patterns in order to detect and identify diseases at an early stage.

Image-based plant phenotyping is a growing application area of computer vision in agriculture. A key task is the segmentation of all individual leaves in images. Here we focus on the most common rosette model plants, Arabidopsis and young tobacco. Although leaves do share appearance and shape characteristics, the presence of occlusions and variability in leaf shape and pose, as well as imaging conditions, render this problem challenging. The aim of this paper is to compare several leaf segmentation solutions on a unique and first-of-its-kind dataset containing images from typical phenotyping experiments. In particular, we report and discuss methods and findings of a collection of submissions for the first Leaf Segmentation Challenge of the Computer Vision Problems in Plant Phenotyping workshop in 2014. Four methods are presented: three segment leaves by processing the distance transform in an unsupervised fashion, and the other via optimal template selection and Chamfer matching. Overall, we find that although separating plant from background can be accomplished with satisfactory accuracy ((Formula presented.)90 % Dice score), individual leaf segmentation and counting remain challenging when leaves overlap. Additionally, accuracy is lower for younger leaves. We find also that variability in datasets does affect outcomes. Our findings motivate further investigations and development of specialized algorithms for this particular application, and that challenges of this form are ideally suited for advancing the state of the art. Data are publicly available (online at http://www.plant-phenotyping.org/datasets) to support future challenges beyond segmentation within this application domain.

High-throughput automated plant phenotyping has recently received a lot of attention. Leaf area is an important characteristic in understanding plant performance, but time-consuming and destructive to measure accurately. In this research, we describe a method to use a histogram of image intensities to automatically measure plant leaf area of tall pepper (Capsicum annuum L.) plants in the greenhouse. With a device equipped with several cameras, images of plants were recorded at 5-cm intervals over a height of 3 m, at a recording distance of less than 60 cm. The images were reduced to a small set of principal components that defined the design matrix in a regression model for predicting manually measured leaf area as obtained from destructive harvesting. These regression calibrations were performed for six different developmental times. In addition, development of leaf area was investigated by fitting linear relations between predicted leaf area and time, with special attention given to the genotype by time interaction and its genetic basis in the form of quantitative trait loci (QTLs). The experiment comprised parents, F1 progeny and eight genotypes of a recombinant inbred population of pepper. Although the current trial contained a limited number of genotypes, an earlier identified QTL related to leaf area growth could be confirmed. Therefore, image analysis, as presented in this paper, provides a powerful and efficient way to study and identify the genetic basis of growth and developmental processes in plants.

In this chapter, an overview of different plant features is given, from (sub)cellular to canopy level. A myriad of methods is available to measure these features using image analysis, and often, multiple methods can be used to measure the same feature. Several criteria are listed for choosing a certain (set of) image descriptor(s) to measure a plant feature. The choice is dependent on a variety of reasons, including accuracy, robustness, recording time, throughput, costs and flexibility. We conclude that hyperspectral imaging can provide a powerful set of image descriptors, which can be used to measure numerous plant features using multivariate statistical models. However, care should be taken that the estimates obtained with these statistical models provide the right measurement for the plant feature under all circumstances of interest.

Plant phenotyping is an emerging science that links genomics with functional plant characteristics. The recent availability of extremely fast high-throughput genotyping technologies has invoked high-throughput phenotyping to become a major bottleneck in the plant breeding programs. As a consequence new camera-based technologies to relieve the phenotyping bottleneck attract considerable attention. Whereas most plant phenotyping technologies are based on the approach to bring the plants to the image recording device, our system brings the camera-system to the plants creating great flexibility on observing plants under practical growing conditions. A new camera, based on lightfield technology was used for image recording. This single lens 3D camera is constructed by placing a micro lens array in front of the image sensor. This also allows the user to change the focus and the point of view after a picture has been taken. While the concept of such plenoptic cameras is known since 1908, only recently the increased computing power of low-cost hardware and the advances in micro lens array production, have made the application of plenoptic cameras feasible. Since this camera outputs a pixel to pixel registered color image and depth map, it solves limitations of common used techniques such as stereo vision and time of flight. During the summer of 2013 an experiment is carried out in a commercial tomato greenhouse in the Netherlands. In this paper first preliminary results are presented and the performance of lightfield technology for plant phenotyping is discussed.

Leaf area measurements are commonly obtained by destructive and laborious practice. This study shows how stereo and time-of-flight (ToF) images can be combined for non-destructive automatic leaf area measurements. The authors focus on some challenging plant images captured in a greenhouse environment, and show that even the state-of-the-art stereo methods produce unsatisfactory results. By transforming depth information in a ToF image to a localised search range for dense stereo, a global optimisation strategy is adopted for producing smooth results that preserve discontinuity. They also use edges of colour and disparity images for automatic leaf detection and develop a smoothing method necessary for accurately estimating surface area. In addition to show that combining stereo and ToF images gives superior qualitative and quantitative results, 149 automatic measurements on leaf area using the authors system in a validation trial have a correlation of 0.97 with true values and the root-mean-square error is 10.97 cm(2), which is 9.3% of the average leaf area. Their approach could potentially be applied for combining other modalities of images with large difference in image resolutions and camera positions.

In our post-genomic world, where we are deluged with genetic information, the bottleneck to scientific progress is often phenotyping, i.e. measuring the observable characteristics of living organisms, such as counting the number of fruits on a plant. Image analysis is one route to automation. In this paper we present a method for recognising and counting fruits from images in cluttered greenhouses. The plants are 3-m high peppers with fruits of complex shapes and varying colours similar to the plant canopy. Our calibration and validation datasets each consist of over 28,000 colour images of over 1000 experimental plants. We describe a new two-step method to locate and count pepper fruits: the first step is to find fruits in a single image using a bag-of-words model, and the second is to aggregate estimates from multiple images using a novel statistical approach to cluster repeated, incomplete observations. We demonstrate that image analysis can potentially yield a good correlation with manual measurement (94.6%) and our proposed method achieves a correlation of 74.2% without any linear adjustment for a large dataset.

Tulip breaking virus (TBV) causes severe economic losses in flower bulbs in the Netherlands. To prevent further spread by aphids, infected plants must be removed from the field as soon as possible. Until now screening is done by visual inspection in the field. As the availability of human experts is limited there is an urgent need for a rapid, automated and objective method of screening. Based on laboratory experiments, we developed a vision method for use in the open field. From 2009 - 2012 field trials were carried out and the techniques were tested and improved. First field trails were on single plant density where plants does not overlap in the images. In 2012 the experiment conducted a tulip field planted at production density of 100 and 125 plants per square meter, resulting in images with overlapping plants. The final score of our system in this production density experiment approached the scores obtained by the experienced crop experts.

Publication Inra prise en compte dans l'analyse bibliométrique des publications scientifiques mondiales sur les Fruits, les Légumes et la Pomme de terre. Période 2000-2012. http://prodinra.inra.fr/record/256699; Most high-throughput systems for automated plant phenotyping involve a fixed recording cabinet to which plants are transported. However, important greenhouse plants like pepper are too tall to be transported. In this research we developed a system to automatically measure plant characteristics of tall pepper plants in the greenhouse. With a device equipped with multiple cameras, images of plants are recorded at a 5 cm interval over a height of 3 m. Two types of features are extracted: (1) features from a 3D reconstruction of the plant canopy; and (2) statistical features derived directly from RGB images. The experiment comprised 151 genotypes of a recombinant inbred population of pepper, to examine the heritability and quantitative trait loci (QTL) of the features. Features extracted from the 3D reconstruction of the canopy were leaf size and leaf angle, with heritabilities of 0.70 and 0.56 respectively. Three QTL were found for leaf size, and one for leaf angle. From the statistical features, plant height showed a good correlation (0.93) with manual measurements, and QTL were in accordance with QTL of manual measurements. For total leaf area, the heritability was 0.55, and two of the three QTL found by manual measurement were found by image analysis.

Tulip breaking virus (TBV) causes severe economic losses in flower bulbs in the Netherlands. To prevent further spread by aphids, the vector of the disease, infected plants must be removed from the field as soon as possible. Until now screening has been carried out by visual inspection in the field. As the availability of human experts is limited there is an urgent need for a rapid, automated and objective method of screening. Based on laboratory experiments, a vision method for use in open fields has been developed. In the period 2009e2012 field trials were carried out and the techniques were tested and improved. During the final evaluation of our system, in the last experiment (2012), the system approached the scores obtained by the experienced crop experts.