Plant Seedlings Dataset

The Plant Seedlings Dataset contains images of approximately 960 unique plants belonging to 12 species at several growth stages.
It comprises annotated RGB images with a physical resolution of roughly 10 pixels per mm.

The database have been recorded at Aarhus University Flakkebjerg Research station in a collaboration between University of Southern Denmark and Aarhus University.

We hope that the database will provide researchers a foundation for training weed recognition algorithms. For more info about dataset, see the dataset description paper

Download links

The dataset contains three files: Full images, automatically segmented plants, and single plants that are not segmented:

Full images

Raw images (9.7GB)

Cropped plants

V2

Some Samples in V1 contained multiple plants. These samples have now been removed.
V2: Nonsegmented single plants (1.7GB)

V1 (used in Kaggle kompetition)

V1: Nonsegmented single plants (1.7GB)

Segmented Cropped plants

Segmented single plants (258MB)

NB: segmentation was made automatically, and should not be considered ground truth.

Content

The database consists of the following species:

Danish	English	Latin	EPPO
Majs	Maize	Zea mays L.	ZEAMX
Vinterhvede	Common wheat	Triticum aestivum L.	TRZAX
Sukkerroe	Sugar beet	Beta vulgaris var. altissima	BEAVA
Lugtløs kamille	Scentless Mayweed	Matricaria perforata Mérat	MATIN
Fuglegræs	Common Chickweed	Stellaria media	STEME
Hyrdetaske	Shepherd’s Purse	Capsella bursa-pastoris	CAPBP
Burresnerre	Cleavers	Galium aparine L.	GALAP
Agersennep	Charlock	Sinapis arvensis L.	SINAR
Hvidmelet gåsefod	Fat Hen	Chenopodium album L.	CHEAL
Liden storkenæb	Small-flowered Cranesbill	Geranium pusillum	GERSS
Agerrævehale	Black-grass	Alopecurus myosuroides	ALOMY
Vindaks	Loose Silky-bent	Apera spica-venti	APESV

Samples

Below you will find an example image taken from the database:

Below you will find samples of each species from the database:

Copyrigth and license

The images and annotations are distributed under the Creative Commons BY-SA license.

All use of the data and derived work, including, but not limited to, trained algorithms and machine learning models requires full citation.

THE IMAGES AND ANNOTATIONS ARE PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS “AS IS” AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS DATA, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

Citation

If you use this dataset in your research or elsewhere, please cite/reference the following paper:
PAPER: A Public Image Database for Benchmark of Plant Seedling Classification Algorithms

Bibtex

@article{Giselsson2017,
author = {Giselsson, Thomas Mosgaard and Dyrmann, Mads and J{\o}rgensen, Rasmus Nyholm and Jensen, Peter Kryger and Midtiby, Henrik Skov},
journal = {arXiv preprint},
keywords = {benchmark,database,plant seedlings,segmentation,site-specific weed control},
title = {{A Public Image Database for Benchmark of Plant Seedling Classification Algorithms}},
year = {2017}
}

Kaggle competition

The images from this dataset have been subject to a Kaggle image-classification competition. We encourage all to take a look at the dataset and commit their solution to the competition. If you are interested in testing your algorithms on weed images ‘from the wild’ with no artificial lighting, you can find some samples at:
Images From The Wild (15MB)

Responsible for the page’s content: Mads Dyrmann