The data set available here is published along with an article “Kraft et al. (2022). Towards operational phytoplankton recognition with automated high-throughput imaging, near real-time data processing, and convolutional neural networks. Front Mar. Sci. 9. Doi: 10.3389/fmars.2022.867695” and if used for further purposes, the article should be cited accordingly. The data set contains approximately 63 000 images belonging to 50 different classes, consisting mainly of phytoplankton. The images can be used to e.g. train a classifier to identify phytoplankton images.

The images were collected with an Imaging FlowCytobot (IFCB, McLane Research Laboratories, Inc., U.S., Olson and Sosik, 2007) from different locations in the Baltic Sea. In 2017 and 2018 the data were collected from a continuous deployment at the Utö Atmospheric and Marine Research Station (59°46.84' N, 21°22.13' E; Laakso et al., 2018; Kraft et al., 2021) operated by Finnish Environment Institute and Finnish Meteorological Institute (n=62). In 2016 and 2019 water samples were collected using the Alg@line ferrybox systems of M/S Finnmaid and Silja Serenade (Ruokanen et al., 2003; Kaitala et al., 2014) and manually ran in the laboratory (n=52). The images were manually annotated by expert taxonomists. The class list and labeled image set is a continuous work in progress, thus there may be a need for revision in future. The data set available with this doi will not be revised. More detailed explanation and example images can be found from the publication Kraft et al. 2022.

The zipped folder contains 50 different folders, and the images are located in the class-specific folders. The image names may refer to an old class (e.g. folder Cryptophyceae-Teleaulax contains images with names Cryptophyceae_drop, Cryptophyceae_small, Teleaulax sp.) that has been joined with another one / revised otherwise.

The work utilized SYKE and FMI marine research infrastructure as a part of the national FINMARI RI consortium. The work was partly funded by Tiina and Antti Herlin Foundation (personal grant for KK), Academy of Finland project FASTVISION (grant no. 321980), Academy of Finland project FASTVISION-plus (grant no. 339355), JERICO-S3 project, funded by the European Commission's H2020 Framework Programme under grant agreement No. 871153, and PHIDIAS project, funded by the European Union's Connecting Europe Facility under grant agreement INEA/CEF/ICT/A2018/1810854.

Original dataset available online at: https://b2share.eudat.eu/records/abf913e5a6ad47e6baa273ae0ed6617a.
Original dataset license: <cc-by-nc-4.0>.

Details

train split means (RGB): [0.6664649575437328]
train split standard deviations (RGB): [0.12305847521607699]

Samples per class for split train

0: Amylax_triacantha                    19.00
1: Aphanizomenon_flosaquae             ▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 6989.00
2: Aphanothece_paralleliformis          29.00
3: Beads                                125.00
4: Centrales_sp                        ▇ 480.00
5: Ceratoneis_closterium                45.00
6: Chaetoceros_sp                      ▇▇▇▇ 1382.00
7: Chaetoceros_sp_single               ▇ 213.00
8: Chlorococcales                       95.00
9: Chroococcales                        142.00
10: Chroococcus_small                  ▇▇ 827.00
11: Ciliata                            ▇ 243.00
12: Cryptomonadales                    ▇▇ 713.00
13: Cryptophyceae-Teleaulax            ▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 6830.00
14: Cyclotella_choctawhatcheeana        102.00
15: Cymbomonas_tetramitiformis         ▇ 199.00
16: Dinophyceae                        ▇▇▇▇ 1433.00
17: Dinophysis_acuminata               ▇ 217.00
18: Dolichospermum-Anabaenopsis        ▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 12280.00
19: Dolichospermum-Anabaenopsis-coiled ▇▇▇▇▇▇▇ 2504.00
20: Euglenophyceae                      102.00
21: Eutreptiella_sp                    ▇▇▇▇▇▇ 2247.00
22: Gonyaulax_verior                    22.00
23: Gymnodiniales                       69.00
24: Gymnodinium_like                    158.00
25: Heterocapsa_rotundata              ▇▇ 614.00
26: Heterocapsa_triquetra              ▇▇▇▇▇▇▇▇▇ 3276.00
27: Heterocyte                         ▇ 263.00
28: Katablepharis_remigera              54.00
29: Licmophora_sp                       74.00
30: Melosira_arctica                    43.00
31: Merismopedia_sp                     98.00
32: Mesodinium_rubrum                  ▇▇▇ 1132.00
33: Monoraphidium_contortum            ▇ 327.00
34: Nitzschia_paleacea                  65.00
35: Nodularia_spumigena                 169.00
36: Oocystis_sp                        ▇▇ 842.00
37: Oscillatoriales                    ▇▇▇▇▇▇▇▇▇▇▇▇ 4440.00
38: Pauliella_taeniata                  119.00
39: Pennales_sp_thick                  ▇ 210.00
40: Pennales_sp_thin                   ▇▇ 781.00
41: Peridiniella_catenata_chain        ▇ 193.00
42: Peridiniella_catenata_single       ▇▇ 899.00
43: Prorocentrum_cordatum              ▇ 276.00
44: Pseudopedinella_sp                 ▇ 379.00
45: Pyramimonas_sp                     ▇▇▇ 1224.00
46: Skeletonema_marinoi                ▇▇▇▇▇▇▇▇▇▇▇ 4128.00
47: Snowella-Woronichinia              ▇▇▇▇▇▇▇▇ 2950.00
48: Thalassiosira_levanderi            ▇▇▇▇▇▇▇ 2537.00
49: Uroglenopsis_sp                    ▇ 516.00

Reference

Kraft, K., Velhonoja, O., Seppälä, J., Hällfors, H., Suikkanen, S., Ylöstalo, P., Anglès, S., Kielosto, S., Kuosa, H., Lehtinen, S., Oja, J., & Tamminen, T. (2022). SYKE-plankton_IFCB_2022 [Data set]. B2SHARE v2. https://doi.org/10.23728/b2share.abf913e5a6ad47e6baa273ae0ed6617a

BibTEX

@dataset{dataset:sykezooscan2024,
  author = {Kraft, Kaisa and Velhonoja, Olli and Seppälä, Jukka and Hällfors, Hanna and Suikkanen, 
            Sanna and Ylöstalo, Petri and Anglès, Susana and Kielosto, Sami and Kuosa, Harri and 
            Lehtinen, Sari and Oja, Jussi and Tamminen, Timo},
  title = {SYKE-plankton_IFCB_2022}, 
  howpublished = {\url{https://doi.org/10.23728/b2share.abf913e5a6ad47e6baa273ae0ed6617a}}, 
  doi = {10.23728/b2share.abf913e5a6ad47e6baa273ae0ed6617a}
  year = {2022}, 
  note = {B2SHARE} 
}

Usage

from datasets import load_dataset

dataset = load_dataset("project-oceania/syke_ifcb_2022")

Downloads last month: 8

Total file size:

495 MB