Annotated image dataset of fire blight symptoms for object detection in orchards.

The monitoring of plant diseases in nurseries, breeding farms and orchards is essential for maintaining plant health. Fire blight (Erwinia amylovora) is still one of the most dangerous diseases in fruit production, as it can spread epidemically and cause enormous economic damage. All measures are therefore aimed at preventing the spread of the pathogen in the orchard and containing an infection at an early stage [1-6]. Efficiency in plant disease control benefits from the development of a digital monitoring system if the spatial and temporal resolution of disease monitoring in orchards can be increased [7]. In this context, a digital disease monitoring system for fire blight based on RGB images was developed for orchards. Between 2021 and 2024, data was collected on nine dates under different weather conditions and with different cameras. The data source locations in Germany were the experimental orchard of the Julius Kühn Institute (JKI), Institute of Plant Protection in Fruit Crops and Viticulture in Dossenheim, the experimental greenhouse of the Julius Kühn Institute for Resistance Research and Stress Tolerance in Quedlinburg and the experimental orchard of the JKI for Breeding Research on Fruit Crops located in Dresden-Pillnitz. The RGB images were taken on different apple genotypes after artificial inoculation with Erwinia amylovora, including cultivars, wild species and progeny from breeding. The presented ERWIAM dataset contains manually labelled RGB images with a size of 1280  × 1280 pixels of fire blight infected shoots, flowers and leaves in different stages of development as well as background images without symptoms. In addition, symptoms of other plant diseases were acquired and integrated into the ERWIAM dataset as a separate class. Each fire blight symptom was annotated with the Computer Vision Annotation Tool (CVAT [8]) using 2-point annotations (bounding boxes) and presented in YOLO 1.1 format (.txt files). The dataset contains a total of 1611 annotated images and 87 background images. This dataset can be used as a resource for researchers and developers working on digital systems for plant disease monitoring.

Rapid and low-cost insect detection for analysing species trapped on yellow sticky traps.

While insect monitoring is a prerequisite for precise decision-making regarding integrated pest management (IPM), it is time- and cost-intensive. Low-cost, time-saving and easy-to-operate tools for automated monitoring will therefore play a key role in increased acceptance and application of IPM in practice. In this study, we tested the differentiation of two whitefly species and their natural enemies trapped on yellow sticky traps (YSTs) via image processing approaches under practical conditions. Using the bag of visual words (BoVW) algorithm, accurate differentiation between both natural enemies and the Trialeurodes vaporariorum and Bemisia tabaci species was possible, whereas the procedure for B. tabaci could not be used to differentiate this species from T. vaporariorum. The decay of species was considered using fresh and aged catches of all the species on the YSTs, and different pooling scenarios were applied to enhance model performance. The best performance was reached when fresh and aged individuals were used together and the whitefly species were pooled into one category for model training. With an independent dataset consisting of photos from the YSTs that were placed in greenhouses and consequently with a naturally occurring species mixture as the background, a differentiation rate of more than 85% was reached for natural enemies and whiteflies.

Sensor-based outdoor monitoring of insects in arable crops for their precise control.

The implementation of precision farming technologies into agricultural practice requires, among other things, precise determination of the extent and intensity of insect infestation in the farmer' fields. Manual insect identification is time-consuming and has low efficiency, especially for large fields. Therefore, scientists and practitioners devote much effort to the automatization of this process. There are two complementary approaches to insect identification: (i) direct, in which the insect (ultimately the species) is determined, and (ii) indirect, in which the damage caused by the insects is monitored and forms the basis on which to formulate the information about insect infestation. A mini-review of both approaches is presented in this work. Additionally, the advantages and disadvantages of each are briefly described. Methods of insect identification are still characterized by relatively small selectivity and efficiency, therefore it is necessary to keep searching for new methods and improve the development of existing ones. The goal of such systems should be to work in real time and be inexpensive to run, enabling widespread use amongst farmers. A possible solution seems to be integrating various techniques (sensor fusion) into a single measurement system. © 2020 Society of Chemical Industry.

An automated field phenotyping pipeline for application in grapevine research.

Due to its perennial nature and size, the acquisition of phenotypic data in grapevine research is almost exclusively restricted to the field and done by visual estimation. This kind of evaluation procedure is limited by time, cost and the subjectivity of records. As a consequence, objectivity, automation and more precision of phenotypic data evaluation are needed to increase the number of samples, manage grapevine repositories, enable genetic research of new phenotypic traits and, therefore, increase the efficiency in plant research. In the present study, an automated field phenotyping pipeline was setup and applied in a plot of genetic resources. The application of the PHENObot allows image acquisition from at least 250 individual grapevines per hour directly in the field without user interaction. Data management is handled by a database (IMAGEdata). The automatic image analysis tool BIVcolor (Berries in Vineyards-color) permitted the collection of precise phenotypic data of two important fruit traits, berry size and color, within a large set of plants. The application of the PHENObot represents an automated tool for high-throughput sampling of image data in the field. The automated analysis of these images facilitates the generation of objective and precise phenotypic data on a larger scale.

Spectrophotometric analyses of chlorophyll and single carotenoids during fruit development of tomato (Solanum lycopersicum L.) by means of iterative multiple linear regression analysis.

When using spectrophotometric transmittance readings of fruit extracts, the analysis of single carotenoids is difficult because of coinciding absorption bands of the various carotenoids and chlorophylls present in the solution. Aimed at the separate analyses of pigments, an iteratively applied linear regression was developed based on spectral profiles of pigment standards. The iterative approach was validated by dilution series of pigments and compared with commonly applied equation systems. High coefficients of determination and low measuring uncertainties were found for chlorophyll a and b (R(2) > or = 0.99, root mean square error RMSE < or = 10%). Carotenoids were separately analyzed with R(2) = 0.99, R(2) = 0.96, and R(2) = 0.98 for lycopene, beta-carotene, and lutein, respectively. The approach based on the spectral profiles provided low measuring uncertainties even if lutein was additionally present in the solutions, which was not possible with common data analyses. Subjecting tomato tissues (Solanum lycopersicum L.) to the iterative approach, contents of in vivo measured pigments were calculated with R(2) = 0.82, R(2) = 0.84, R(2) = 0.67, and R(2) = 0.03 for chlorophyll a and b, lycopene, and beta-carotene, respectively.