Annotated dataset of microscope images of pollen grains from 40 beekeeping taxa
Creators
Description
The study of beekeeping flora and the analysis and identification of pollen collected by bees are important tools for beekeepers and researchers seeking to understand bee feeding habits and assess the ecological interactions between bees and plants. To identify the botanical origin of the pollen pellets collected by the bees, palynology method is mainly followed.
Pollen grains show great diversity, in terms of size, shape, symmetry and surface, as well as in terms of the number and type of their openings (apertures). They often present openings on their outer surface, serving as excellent diagnostic characters, as they show stability in their form and number. The most common types of openings are the pores, the sinuses, and the combination of the above, the anal canals. Pollen grains with pores are characterized as porate, those with sinuses as colpate, if they contain both as colporate. Depending on the number of openings, corresponding prefixes such as mono-, di-, tri- etc. precede the above terms. Also, the position of the openings, whether they are at the poles or in the equatorial zone, as well as their shape, are taken also into account.
Pollen size can be used as a diagnostic feature, but it shows great variability even within the same pollen grain. This is because it can be affected by various factors such as chemical treatment and the materials used in the creation of the preparations, genetic variability, etc. Specifically, a frequent phenomenon is the shrinking of pollen grains (harmomegathy or Wodehouse effect) resulting from the change in the bursting pressure of the cytoplasm during the hydration or dehydration of the pollen. Therefore, the degree of hydration is responsible for the actual shape and size of the pollen grains.
Considering all the above, a database was created including microscope images and characteristics (such as the type of pollen grain, the size, the type and number of openings, etc) of 40 taxa of major beekeeping importance.
Pollen in the form of pellets was crushed and a small amount was taken with special stainless steel forceps and then mixed with a drop of 20% glucose solution on a slide. Fuchsin solution was added and the preparation was spread over a 22 x 22 mm surface. The preparations were dried by gentle heating to 40°C, on a heating plate and a cover slip is placed to which a small amount of Entellan was added. All pollen grains were photographed on an optical microscope (Olympus SZX12), with lens 40× (Olympus DF PLAPO 1X DF), with a digital analysis camera (Olympus SC30), while a morphometry software (Image Pro Plus Software, V1.1.19) was used for their determination. For the microscopic identification of the pollen types, the collection of reference slides from the Laboratory of Apiculture of the Aristotle University of Thessaloniki, which is accredited to ISO 17025:2017, was used.
The dataset contains 3379 training captured microscope images of pollen grains from 40 major beekeeping taxa (class list can be found below) and 85 testing captured images. Polygon annotations (files train.json and val.json included) were created using LabelMe software and saved in COCO Annotation format.
Further information about the related project (SmartBeeKeep) can be found in the following article and presentation (please site if you use these data):
- Vasilios Liolios, Dimitrios Kanelis, Maria-Anna Rodopoulou, Chrysoula Tananaki (2023). A Comparative Study of Methods Recording Beekeeping Flora. Forests, 14(8), 1677; https://doi.org/10.3390/f14081677
- Nikos Grammalidis, Andreas Stergioulas, Aggelos Avramidis, Konstantinos Karystinakis, Athanasios Partozis, Athanasios Topaloudis, Georgia Kalantzi, Chrisoula Tananaki, Dimitrios Kanelis, Vasilis Liolios, and Madesis Panagiotis "A smart beekeeping platform based on remote sensing and artificial intelligence", Proc. SPIE 12786, Ninth International Conference on Remote Sensing and Geoinformation of the Environment (RSCy2023), 127860C (21 September 2023); https://doi.org/10.1117/12.2681866 Event: Ninth International Conference on Remote Sensing and Geoinformation of the Environment (RSCy2023), 2023, Ayia Napa, Cyprus Author preprint available
Annotation - Latin name
Anthemis - Anthemis sp.
Asphodelus - Asphodelus fistulosus
Brassica napus - Brassica napus
Castanea - Castanea sativa
Cephalaria - Cephalaria transsylvanica
Chenopodium - Chenopodium album
Cichorium intybus - Cichorium intybus
Cistus - Cistus creticus
Cistus salvifolius - Cistus salvifolius
Convolvulus - Convolvulus arvensis
Daucus - Daucus carota
Echium - Echium plantagineum
Erica - Erica manipuliflora
Hederahelix - Hedera helix
Helianthus - Helianthus annuus
Heliotropium - Heliotropium europaeum
Hypericum - Hypericum perforatum
Lavandula - Lavandula angustifolia
Ligustrum - Ligustrum japonicum
Matricaria - Matricaria chamomilla
Olea - Olea europaea
Paliurus - Paliurus spina-christi
Papaver - Papaver rhoeas
Pinus - Pinus sp.
Polugonum - Polygonum aviculare
Portulaca - Portulaca oleracea
Pyrus spinosa - Pyrus spinosa
Quercussp - Quercus coccifera
Rosmarinus officinalis - Rosmarinus officinalis
Rubus - Rubus ulmifolius
Silybum marianum - Silybum marianum
Sinapis - Sinapis arvensis
sonchus - Sonchus asper
Taraxacum officinale - Taraxacum officinale
Tamarix - Tamarix sp.
Tilia intermedia - Tilia sp.
Tribulus - Tribulus terrestis
Trifolium - Trifolium campestre
Verbascum - Verbascum nigrum
Vicia vilosa - Vicia villosa
Files
Pollen_grains_40_classes.zip
Files
(746.3 MB)
Name | Size | Download all |
---|---|---|
md5:91a365b8294f6f077530bf8babc3963a
|
746.3 MB | Preview Download |