Introduction #
The authors of the Severstal: Steel Defect Detection dataset acknowledge that steel holds a paramount position as one of the most vital building materials in modern construction. Its resilience against both natural elements and wear caused by human activities has rendered it indispensable worldwide. In the pursuit of enhancing the efficiency of steel production, the Severstal competition aims to play a pivotal role in the detection of defects within the steel production process.
The intricate process of producing flat sheet steel demands precision and care at every step, from heating and rolling to drying and cutting. Numerous machines come into contact with flat steel before it’s deemed ready for use. At present, Severstal employs images from high-frequency cameras to drive a defect detection algorithm.
In this competition, participants are tasked with aiding engineers in refining this algorithm by localizing and categorizing surface defects on steel sheets.
This competition aims to predict the location and type of defects found in steel manufacturing by segmenting and classifying the defects in the test set.
Each image may have no defects, a defect of a single class, or defects of multiple classes. There are 3 different defect classes.
Summary #
Severstal: Steel Defect Detection 2019 Challenge is a dataset for instance segmentation, semantic segmentation, and object detection tasks. It is used in the surface defect detection domain, and in the manufacturing industry.
The dataset consists of 18074 images with 19958 labeled objects belonging to 4 different classes including defect_3, defect_1, defect_4, and other: defect_2.
Images in the Severstal dataset have pixel-level instance segmentation annotations. Due to the nature of the instance segmentation task, it can be automatically transformed into a semantic segmentation (only one mask for every class) or object detection (bounding boxes for every object) tasks. There are 11408 (63% of the total) unlabeled images (i.e. without annotations). There are 2 splits in the dataset: train (12568 images) and test (5506 images). The dataset was released in 2019 by the Severstal, Russia.
Here is the visualized example grid with animated annotations:
Explore #
Severstal dataset has 18074 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.
Class balance #
There are 4 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.
Class ã…¤ | Images ã…¤ | Objects ã…¤ | Count on image average | Area on image average |
---|---|---|---|---|
defect_3âž” mask | 5150 | 14648 | 2.84 | 6.22% |
defect_1âž” mask | 897 | 3082 | 3.44 | 1.06% |
defect_4âž” mask | 801 | 1907 | 2.38 | 8.39% |
defect_2âž” mask | 247 | 321 | 1.3 | 0.82% |
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.
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 |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
defect_3 mask | 14648 | 2.19% | 80.53% | 0% | 1px | 0.39% | 256px | 100% | 155px | 60.74% | 1px | 0.06% | 1600px | 100% |
defect_1 mask | 3082 | 0.31% | 4.26% | 0.02% | 13px | 5.08% | 256px | 100% | 58px | 22.55% | 6px | 0.38% | 180px | 11.25% |
defect_4 mask | 1907 | 3.52% | 41.05% | 0% | 1px | 0.39% | 256px | 100% | 113px | 44% | 1px | 0.06% | 1234px | 77.12% |
defect_2 mask | 321 | 0.64% | 2.35% | 0.05% | 48px | 18.75% | 256px | 100% | 171px | 66.66% | 6px | 0.38% | 45px | 2.81% |
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 19958 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.
Object ID ã…¤ | Class ã…¤ | Image name click row to open | Image size height x width | Height ã…¤ | Height ã…¤ | Width ã…¤ | Width ã…¤ | Area ã…¤ |
---|---|---|---|---|---|---|---|---|
1âž” | defect_3 mask | 7378f1d94.jpg | 256 x 1600 | 41px | 16.02% | 44px | 2.75% | 0.32% |
2âž” | defect_3 mask | 7378f1d94.jpg | 256 x 1600 | 39px | 15.23% | 25px | 1.56% | 0.17% |
3âž” | defect_3 mask | 7378f1d94.jpg | 256 x 1600 | 60px | 23.44% | 40px | 2.5% | 0.46% |
4âž” | defect_3 mask | 8b9681587.jpg | 256 x 1600 | 24px | 9.38% | 58px | 3.62% | 0.27% |
5âž” | defect_3 mask | 8b9681587.jpg | 256 x 1600 | 72px | 28.12% | 118px | 7.38% | 1.2% |
6âž” | defect_3 mask | 8b9681587.jpg | 256 x 1600 | 34px | 13.28% | 64px | 4% | 0.43% |
7âž” | defect_1 mask | 3fdc95c42.jpg | 256 x 1600 | 70px | 27.34% | 40px | 2.5% | 0.57% |
8âž” | defect_1 mask | 3fdc95c42.jpg | 256 x 1600 | 63px | 24.61% | 27px | 1.69% | 0.33% |
9âž” | defect_1 mask | 3fdc95c42.jpg | 256 x 1600 | 32px | 12.5% | 22px | 1.38% | 0.15% |
10âž” | defect_1 mask | 3fdc95c42.jpg | 256 x 1600 | 58px | 22.66% | 29px | 1.81% | 0.32% |
License #
The dataset is owned by PAO Severstal. If you need to use the dataset in your research you are entitled to do it without any restrictions except those established by the applicable law.
Citation #
If you make use of the Severstal data, please cite the following reference:
@misc{severstal-steel-defect-detection,
author = {Alexey Grishin, BorisV, iBardintsev, inversion, Oleg},
title = {Severstal: Steel Defect Detection},
publisher = {Kaggle},
year = {2019},
url = {https://kaggle.com/competitions/severstal-steel-defect-detection}
}
If you are happy with Dataset Ninja and use provided visualizations and tools in your work, please cite us:
@misc{ visualization-tools-for-severstal-dataset,
title = { Visualization Tools for Severstal Dataset },
type = { Computer Vision Tools },
author = { Dataset Ninja },
howpublished = { \url{ https://datasetninja.com/severstal } },
url = { https://datasetninja.com/severstal },
journal = { Dataset Ninja },
publisher = { Dataset Ninja },
year = { 2024 },
month = { oct },
note = { visited on 2024-10-15 },
}
Download #
Dataset Severstal 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='Severstal', 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.