AG20000/Agro_Dataset_344k

The Agro_Dataset_344k was created by aggregating 24 publicly available weed (an plant) classification and detection datasets (see refs below) and converted to classification format by cropping bounding boxes.

Dataset Contents 97 Species, 344,556 Images:

Zea mays 29,635 images (8.60%) Veronica arvensis 20,742 images (6.02%) Papaver rhoeas 17,723 images (5.14%) Beta vulgaris 15,985 images (4.64%) Capsella bursa-pastoris 13,600 images (3.95%) Tripleurospermum inodorum 12,342 images (3.58%) Anagallis arvensis 10,107 images (2.93%) Veronica persica 9,103 images (2.64%) Stellaria media 8,245 images (2.39%) Poa annua 7,414 images (2.15%) Urtica urens 7,404 images (2.15%) Chenopodium album 7,108 images (2.06%) Matricaria chamomilla 7,095 images (2.06%) Solanum nigrum 6,316 images (1.83%) Senecio vulgaris 6,256 images (1.82%) Thlaspi arvense 6,235 images (1.81%) Viola arvensis 6,123 images (1.78%) Silene noctiflora 5,660 images (1.64%) Calotropis procera 5,640 images (1.64%) Manihot esculenta 5,569 images (1.62%) Plantago major 4,754 images (1.38%) Sonchus oleraceus 4,660 images (1.35%) Artemisia vulgaris 4,425 images (1.28%) Erodium cicutarium 4,256 images (1.24%) Centaurea cyanus 4,135 images (1.20%) Phaseolus vulgaris 4,108 images (1.19%) Brassica napus 4,032 images (1.17%) Sinapis arvensis 4,019 images (1.17%) Vicia hirsuta 3,914 images (1.14%) Geranium molle 3,592 images (1.04%) Lolium multiflorum 3,473 images (1.01%) Plantago lanceolata 3,416 images (0.99%) Persicaria lapathifolia 3,375 images (0.98%) Myosotis arvensis 3,051 images (0.89%) Amaranthus palmeri 3,050 images (0.89%) Euphorbia geniculata 3,002 images (0.87%) Persicaria maculosa 2,994 images (0.87%) Ipomoea purpurea 2,942 images (0.85%) Avena fatua 2,903 images (0.84%) Apera spica-venti 2,659 images (0.77%) Alopecurus myosuroides 2,471 images (0.72%) Parthenium hysterophorus 2,433 images (0.71%) Amaranthus tuberculatus 2,410 images (0.70%) Polygonum aviculare 2,407 images (0.70%) Mollugo verticillata 2,326 images (0.68%) Gossypium hirsutum 2,324 images (0.67%) Malva parviflora 2,100 images (0.61%) Cucurbita pepo 1,999 images (0.58%) Raphanus raphanistrum 1,914 images (0.56%) Commelina benghalensis 1,704 images (0.49%) Digitaria sanguinalis 1,666 images (0.48%) Bromus sterilis 1,648 images (0.48%) Ipomoea obscura 1,637 images (0.48%) Fallopia convolvulus 1,602 images (0.46%) Galium aparine 1,557 images (0.45%) Cirsium arvense 1,502 images (0.44%) Portulaca oleracea 1,442 images (0.42%) Lolium rigidum 1,354 images (0.39%) Lapsana communis 1,338 images (0.39%) Cyperus rotundus 1,316 images (0.38%) Chamaecrista pumila 1,306 images (0.38%) Glebionis segetum 1,277 images (0.37%) Convolvulus arvensis 1,258 images (0.37%) Euphorbia maculata 1,186 images (0.34%) Ziziphus mauritiana 1,125 images (0.33%) Glycine max 1,124 images (0.33%) Senna obtusifolia 1,124 images (0.33%) Eclipta prostrata 1,119 images (0.32%) Sesamum indicum 1,095 images (0.32%) Chromolaena odorata 1,074 images (0.31%) Lantana camara 1,064 images (0.31%) Vachellia nilotica 1,062 images (0.31%) Amsinckia 1,041 images (0.30%) Parkinsonia aculeata 1,031 images (0.30%) Ambrosia artemisiifolia 1,025 images (0.30%) Gutierrezia sarothrae 1,016 images (0.29%) Cryptostegia grandiflora 1,009 images (0.29%) Lolium perenne 921 images (0.27%) Argemone mexicana 854 images (0.25%) Cynodon dactylon 709 images (0.21%) Clitoria ternatea 678 images (0.20%) Sida spinosa 613 images (0.18%) Euphorbia hirta 571 images (0.17%) Boerhaavia diffusa 539 images (0.16%) Taraxacum officinale 531 images (0.15%) Eleusine indica 425 images (0.12%) Anchusa arvensis 379 images (0.11%) Euphorbia hypericifolia 262 images (0.08%) Solanum lycopersicum 201 images (0.06%) Euphorbia helioscopia 172 images (0.05%) Fumaria officinalis 143 images (0.04%) Physalis angulata 116 images (0.03%) Rapistrum rugosum 111 images (0.03%) Lepidium didymum 72 images (0.02%) Anoda cristata 61 images (0.02%) Erigeron bonariensis 30 images (0.01%) Senecio madagascariensis 20 images (0.01%)

The authors gratefully acknowledge the researchers and institutions whose publicly available datasets were used in the aggregation of the Agro Dataset.

References:

Chen, D., Lu, Y., Li, Z., & Young, S. (2022). Performance evaluation of deep transfer learning on multi-class identification of common weed species in cotton production systems. Computers and Electronics in Agriculture, 198, 107091. https://doi.org/10.1016/j.compag.2022.107091

Rashid, M. R. A., Biswas, J., & Hossain, M. M. (2024). Pumpkin leaf diseases dataset from Bangladesh [Data set]. Mendeley Data, V1. https://doi.org/10.17632/wtxcw8wpxb.1

AUA Group. (n.d.). early-crop-weed [GitHub repository]. https://github.com/AUAgroup/early-crop-weed

Rajput, A. S., Shukla, S., & Thakur, S. S. (2023). SoyNet: A high-resolution Indian soybean image dataset for leaf disease classification. Data in Brief, 49, 109447. https://doi.org/10.1016/j.dib.2023.109447

Iqbal, N., Manss, C., Scholz, C., Koenig, D., Igelbrink, M., & Ruckelshausen, A. (2023). AI-based maize and weeds detection on the edge with CornWeed dataset. 18th Conference on Computer Science and Intelligence Systems (FedCSIS). https://doi.org/10.15439/2023F2125

Lu, Y. (2023). CottonWeedDet12: A 12-class weed dataset of cotton production systems for benchmarking AI models for weed detection [Data set]. Zenodo. https://doi.org/10.5281/zenodo.7535814

Lu, Y. (2022). CottonWeedDet3 [Data set]. Kaggle. https://doi.org/10.34740/KAGGLE/DSV/4090494

Salazar-Gomez, A., Darbyshire, M., Gao, J., Sklar, E. I., & Parsons, S. (2021). Towards practical object detection for weed spraying in precision agriculture. arXiv. https://doi.org/10.48550/arXiv.2109.11048

Madsen, S. L., Mathiassen, S. K., Dyrmann, M., Laursen, M. S., Paz, L.-C., & Jørgensen, R. N. (2020). Open plant phenotype database of common weeds in Denmark. Remote Sensing, 12(8), 1246. https://doi.org/10.3390/rs12081246

Plant, B. (2023). Amsinckia in chickpeas [Data set]. Weed-AI Sydney. https://weed-ai.sydney.edu.au/datasets/21675efe-9d25-4096-be76-3a541475efd4

Coleman, G. (2022). Broadleaf weeds in common couch [Data set]. Weed-AI Sydney. https://weed-ai.sydney.edu.au/datasets/8b14a44b-bc7f-4b92-9bc0-224a2a2c4e22

Coleman, G. (2022). Brownlow Hill fireweed [Data set]. Weed-AI Sydney. https://weed-ai.sydney.edu.au/datasets/24b34712-c31b-4efc-9790-406d1f14d840

Coleman, G. (2021). 20200827 — Cobbity wheat BFLY [Data set]. Weed-AI Sydney. https://weed-ai.sydney.edu.au/datasets/3c363da3-6274-45e4-a0ce-b307cb0f89cc

Acharya, R. (2020). Corn leaf infection dataset [Data set]. Kaggle. https://www.kaggle.com/qramkrishna/corn-leaf-infection-dataset

Coleman, G. (2021). 20190728 — Narrabri wheat [Data set]. Weed-AI Sydney. https://weed-ai.sydney.edu.au/datasets/dc322d80-be00-49cf-822c-9e9b40e37425

Coleman, G. (2021). 20200701 — Narrabri chickpea BFLYS [Data set]. Weed-AI Sydney. https://weed-ai.sydney.edu.au/datasets/839a5f35-9c7b-4df3-92f4-d0fc15120920

Coleman, G., Kutugata, M., Walsh, M., & Bagavathiannan, M. (2023). Palmer amaranth growth stage — 8 (PAGS8) [Data set]. Weed-AI Sydney. https://weed-ai.sydney.edu.au/datasets/5c78d067-8750-4803-9cbe-57df8fae55e4

Rayner, G. (2022). RadishWheatDataset [Data set]. Weed-AI Sydney. https://weed-ai.sydney.edu.au/datasets/8b8f134f-ede4-4792-b1f7-d38fc05d8127

Coleman, G. (2021). 20210317 — Wild radish in wheat [Data set]. Weed-AI Sydney. https://weed-ai.sydney.edu.au/datasets/09af32ad-2e9e-4f7c-ae08-55374824ee15

Innovatiana. (n.d.). Crop and weed detection data with bounding boxes [Data set]. https://www.innovatiana.com/en/datasets/crop-and-weed-detection-data-with-bounding-boxes

Shinde, S., & Attar, V. (2024). MH-Weed16: An Indian multiclass annotated weed dataset for computer vision tasks [Data set]. Mendeley Data, V1. https://doi.org/10.17632/d3n3mgjjbv.1

Bertoglio, R., Kalouguine, A., Boffety, D., Boulet, M., Berducat, M., Facchinetti, D., Fontana, G., & Matteucci, M. (n.d.). The ACRE crop-weed dataset [Data set]. Zenodo. https://doi.org/10.5281/zenodo.8102216