DiaMOS Plant - Dataset Ninja

Introduction #

Released 2021-10-08 ·Gianni Fenu, Francesca Maridina Malloci

The authors of the DiaMOS Plant: A Dataset for Diagnosis and Monitoring Plant Disease contribute to the evolving field of foliar disease classification and recognition through the utilization of machine and deep learning concepts. It has 3505 images of pear fruit and leaves affected by four diseases: slug leaf, spot leaf, curl leaf, and healthy leaf. The study offers valuable guidelines for the research community to select and construct further datasets.

The direct visual examination of leaves serves as a crucial source of information for assessing plant health. Leaf symptoms are early indicators of various diseases, infections, parasites, and deficiencies that impact plant development and life cycles. Biotic and abiotic stresses, significant factors limiting agricultural productivity, necessitate innovative and sustainable cultivation practices.

DiaMOS Plant is introduced as a field dataset for diagnosing and monitoring plant symptoms. It covers an entire growing season of a pear tree, containing 3505 images depicting four leaf stresses and three stages of fruit development. The dataset, suitable for machine and deep learning methods, includes detailed information on fruit and leaf images, resolution variations, and acquisition devices.

DiaMOS Plant Dataset
Plant	Pear
Cultivar	Septoria Piricola
Data Source Location	Sardegna, Italy
Type of Data	RGB Images
ROI (Region of Interest)	Leaf, Fruit
Total Size	3505 images

Dataset description

The authors of the dataset comprise images captured using different devices, including smartphones and DSLR cameras. Two resolutions, 2976 × 3968 and 3456 × 5184, add complexity and value to the dataset. Multiple devices were employed to simulate real-world scenarios where operators have diverse technical characteristics in their devices. Leaves were captured from the adaxial side under various realistic conditions, including different lighting, angles, backgrounds, and noise levels. This approach allowed for the representation of leaves under diverse lighting conditions and the observation of symptom evolution over time.

	Smartphone Camera	DSRL Camera
Image size	2976 × 3968	3456 × 5184
Model device	Honor 6×	Canon EOS 60D
Focal length	3.83 mm	50 mm
Focal ratio	f/2.2	f/4.5
Color space	RGB	RGB

Dataset annotation

The dataset labeling process involved manual annotation using the LabelImg software. An expert assisted in disease recognition, and severity levels were assigned for specific classes. Severity levels ranged from no risk (0%) to high risk (>50%), providing detailed information on the affected leaf area.
The dataset’s representativeness, complexity, and detailed annotations contribute to its potential for advancing research in the field of plant symptom diagnosis and monitoring. Some annotations are of poor quality: the bounding boxes do not correspond to the location of objects in the image.

Expand

Homepage

Research Paper 1 (main)

Research Paper 2

GitHub

Summary #

DiaMOS Plant: A Dataset for Diagnosis and Monitoring Plant Disease is a dataset for an object detection task. It is used in the agricultural industry.

The dataset consists of 3505 images with 3753 labeled objects belonging to 5 different classes including slug leaf, spot leaf, pear, and other: curl leaf and healthy leaf.

Images in the DiaMOS Plant dataset have bounding box annotations. There are 13 (0% of the total) unlabeled images (i.e. without annotations). There are no pre-defined train/val/test splits in the dataset. Additionally, the images have severity level and source tags. The severity level is determined based on the percentage of leaf area affected, source - identifies the device with which the image was taken. The dataset was released in 2021 by the University of Cagliari, Italy.

Explore #

DiaMOS Plant dataset has 3505 images. Click on one of the examples below or open "Explore" tool anytime you need to view dataset images with annotations. This tool has extended visualization capabilities like zoom, translation, objects table, custom filters and more. Hover the mouse over the images to hide or show annotations.

Sample annotation mask from DiaMOS Plant

👀

Have a look at 3505 images

View images along with annotations and tags, search and filter by various parameters

Class balance #

There are 5 annotation classes in the dataset. Find the general statistics and balances for every class in the table below. Click any row to preview images that have labels of the selected class. Sort by column to find the most rare or prevalent classes.

Rows 1-5 of 5

Class ㅤ	Images ㅤ	Objects ㅤ	Count on image average	Area on image average
slug leaf➔ rectangle	2025	2028	1	26.69%
spot leaf➔ rectangle	884	885	1	26.73%
pear➔ rectangle	486	734	1.51	14.73%
curl leaf➔ rectangle	54	63	1.17	20.28%
healthy leaf➔ rectangle	43	43	1	27.87%

Co-occurrence matrix #

Co-occurrence matrix is an extremely valuable tool that shows you the images for every pair of classes: how many images have objects of both classes at the same time. If you click any cell, you will see those images. We added the tooltip with an explanation for every cell for your convenience, just hover the mouse over a cell to preview the description.

Images #

Explore every single image in the dataset with respect to the number of annotations of each class it has. Click a row to preview selected image. Sort by any column to find anomalies and edge cases. Use horizontal scroll if the table has many columns for a large number of classes in the dataset.

Object distribution #

Interactive heatmap chart for every class with object distribution shows how many images are in the dataset with a certain number of objects of a specific class. Users can click cell and see the list of all corresponding images.

Class sizes #

The table below gives various size properties of objects for every class. Click a row to see the image with annotations of the selected class. Sort columns to find classes with the smallest or largest objects or understand the size differences between classes.

Rows 1-5 of 5

Class	Object count	Avg area	Max area	Min area	Min height	Min height	Max height	Max height	Avg height	Avg height	Min width	Min width	Max width	Max width
slug leaf rectangle	2028	26.7%	55.62%	8.01%	864px	26.94%	3343px	84.25%	2009px	52.27%	732px	24.6%	3082px	91.7%
spot leaf rectangle	885	26.73%	73.33%	4.12%	836px	24.14%	3773px	95.09%	2077px	53.09%	582px	14.67%	3882px	77.45%
pear rectangle	734	10.29%	31.07%	0%	7px	0.18%	3069px	74.71%	1326px	37.27%	27px	0.91%	2992px	59.88%
curl leaf rectangle	63	17.47%	44.51%	3.76%	574px	16.61%	4160px	80.25%	1860px	46.4%	607px	11.98%	3292px	63.69%
healthy leaf rectangle	43	27.87%	58.77%	9.51%	1275px	33.87%	3122px	79.64%	2144px	55.38%	811px	27.25%	2597px	74.7%

Spatial Heatmap #

The heatmaps below give the spatial distributions of all objects for every class. These visualizations provide insights into the most probable and rare object locations on the image. It helps analyze objects' placements in a dataset.

Objects #

Table contains all 3753 objects. Click a row to preview an image with annotations, and use search or pagination to navigate. Sort columns to find outliers in the dataset.

Rows 1-10 of 3753

Object ID ㅤ	Class ㅤ	Image name click row to open	Image size height x width	Height ㅤ	Height ㅤ	Width ㅤ	Width ㅤ	Area ㅤ
1➔	slug leaf rectangle	u1711.jpg	3968 x 2976	2401px	60.51%	1666px	55.98%	33.87%
2➔	curl leaf rectangle	u3117.jpg	3456 x 5184	1775px	51.36%	1828px	35.26%	18.11%
3➔	slug leaf rectangle	u1623.jpg	3968 x 2976	1801px	45.39%	1549px	52.05%	23.62%
4➔	slug leaf rectangle	u1519.jpg	3968 x 2976	1891px	47.66%	1406px	47.24%	22.52%
5➔	slug leaf rectangle	u1478.jpg	3968 x 2976	2487px	62.68%	1220px	40.99%	25.69%
6➔	slug leaf rectangle	u2286.jpg	3968 x 2976	2340px	58.97%	1657px	55.68%	32.83%
7➔	pear rectangle	234.jpg	3456 x 5184	2137px	61.83%	1129px	21.78%	13.47%
8➔	slug leaf rectangle	u2417.jpg	3968 x 2976	2375px	59.85%	1641px	55.14%	33%
9➔	slug leaf rectangle	u1782.jpg	3968 x 2976	2367px	59.65%	1224px	41.13%	24.53%
10➔	spot leaf rectangle	u2118.jpg	3968 x 2976	2406px	60.64%	1453px	48.82%	29.6%

License #

A Dataset for Diagnosis and Monitoring Plant Disease is under CC BY 4.0 license.

Source

Citation #

If you make use of the DiaMOS Plant data, please cite the following reference:

@dataset{gianni_fenu_2021_5557313,
  author       = {Gianni Fenu and
                  Francesca Maridina Malloci},
  title        = {{DiaMOS Plant Dataset: A Dataset for Diagnosis and
                   Monitoring Plant Disease}},
  month        = oct,
  year         = 2021,
  publisher    = {Zenodo},
  version      = 1,
  doi          = {10.5281/zenodo.5557313},
  url          = {https://doi.org/10.5281/zenodo.5557313}
}

Source

If you are happy with Dataset Ninja and use provided visualizations and tools in your work, please cite us:

@misc{ visualization-tools-for-diamos-plant-dataset,
  title = { Visualization Tools for DiaMOS Plant Dataset },
  type = { Computer Vision Tools },
  author = { Dataset Ninja },
  howpublished = { \url{ https://datasetninja.com/diamos-plant } },
  url = { https://datasetninja.com/diamos-plant },
  journal = { Dataset Ninja },
  publisher = { Dataset Ninja },
  year = { 2024 },
  month = { jul },
  note = { visited on 2024-07-27 },
}

Download #

Dataset DiaMOS Plant can be downloaded in Supervisely format:

As an alternative, it can be downloaded with dataset-tools package:

pip install --upgrade dataset-tools

… using following python code:

import dataset_tools as dtools

dtools.download(dataset='DiaMOS Plant', dst_dir='~/dataset-ninja/')

Make sure not to overlook the python code example available on the Supervisely Developer Portal. It will give you a clear idea of how to effortlessly work with the downloaded dataset.

The data in original format can be downloaded here:

. . .

Disclaimer #

Our gal from the legal dep told us we need to post this:

Dataset Ninja provides visualizations and statistics for some datasets that can be found online and can be downloaded by general audience. Dataset Ninja is not a dataset hosting platform and can only be used for informational purposes. The platform does not claim any rights for the original content, including images, videos, annotations and descriptions. Joint publishing is prohibited.

You take full responsibility when you use datasets presented at Dataset Ninja, as well as other information, including visualizations and statistics we provide. You are in charge of compliance with any dataset license and all other permissions. You are required to navigate datasets homepage and make sure that you can use it. In case of any questions, get in touch with us at hello@datasetninja.com.