Find research datasets worth reusing
Search datasets from major research repositories and use ShareScore to quickly assess how well each record supports discovery, access, and reuse.
8
datasets available to search
ShareScore release 0.9.0
Dataset results
8 results for “road surface”
StreetSurfaceVis: a dataset of street-level imagery with annotations of road surface type and quality
<h1>StreetSurfaceVis</h1> <p><em>StreetSurfaceVis</em> is an image dataset containing <strong>9,122 street-level images from Germany</strong> with labels on <strong>road surface type and quality.</strong> The CSV file <code>streetSurfaceVis_v1_0.csv</code> contains all image metadata and four folders contain the image files. All images are available in four different sizes, based on the image width, in 256px, 1024px, 2048px and the original size.<br>Folders containing the images are named according to the respective image size. Image files are named based on the <code>mapillary_image_id</code>.</p> <p>You can find the corresponding publication here: <a href="https://www.nature.com/articles/s41597-024-04295-9#citeas">StreetSurfaceVis: a dataset of crowdsourced street-level imagery with semi-automated annotations of road surface type and quality</a></p> <p> </p> <h3>Image metadata</h3> <p>Each CSV record contains information about one street-level image with the following attributes:</p> <ul> <li><code>mapillary_image_id</code>: ID provided by Mapillary (see information below on Mapillary)</li> <li><code>user_id</code>: Mapillary user ID of contributor</li> <li><code>user_name</code>: Mapillary user name of contributor</li> <li><code>captured_at</code>: timestamp, capture time of image</li> <li><code>longitude</code>, <code>latitude</code>: location the image was taken at</li> <li><code>train</code>: Suggestion to split train and test data. `True` for train data and `False` for test data. Test data contains data from 5 cities which are excluded in the training data.</li> <li><code>surface_type</code>: Surface type of the road in the focal area (the center of the lower image half) of the image. Possible values: asphalt, concrete, paving_stones, sett, unpaved</li> <li><code>surface_quality</code>: Surface quality of the road in the focal area of the image. Possible values: (1) excellent, (2) good, (3) intermediate, (4) bad, (5) very bad (see the attached <strong>Labeling Guide document</strong> for details)</li> </ul> <p> </p> <h3>Image source</h3> <p>Images are obtained from <a href="https://www.mapillary.com/">Mapillary</a>, a crowd-sourcing plattform for street-level imagery. More metadata about each image can be obtained via the <a href="https://www.mapillary.com/developer/api-documentation">Mapillary API . </a>User-generated images are shared by Mapillary under the <a href="https://creativecommons.org/licenses/by-sa/4.0/">CC-BY-SA</a> License.</p> <p>For each image, the dataset contains the <code>mapillary_image_id</code> and <code>user_name</code>. <br>You can access user information on the Mapillary website by <code>https://www.mapillary.com/app/user/<USER_NAME> </code><br>and image information by <code>https://www.mapillary.com/app/?focus=photo&pKey=<MAPILLARY_IMAGE_ID></code></p> <p>If you use the provided images, please adhere to the <a href="https://www.mapillary.com/terms">terms of use of Mapillary.</a></p> <p> </p> <h3>Instances per class</h3> <p>Total number of images: 9,122</p> <table> <tbody> <tr> <td> </td> <td><strong>excellent</strong></td> <td><strong>good</strong></td> <td><strong>intermediate</strong></td> <td><strong>bad</strong></td> <td><strong>very bad</strong></td> </tr> <tr> <td><strong>asphalt</strong></td> <td>971</td> <td>1697</td> <td>821</td> <td>246</td> <td>-</td> </tr> <tr> <td><strong>concrete</strong></td> <td>314</td> <td>350</td> <td>250</td> <td>58</td> <td>-</td> </tr> <tr> <td><strong>paving stones</strong></td> <td>385</td> <td>1063</td> <td>519</td> <td>70</td> <td>-</td> </tr> <tr> <td><strong>sett</strong></td> <td>-</td> <td>129</td> <td>694</td> <td>540</td> <td>-</td> </tr> <tr> <td><strong>unpaved</strong></td> <td>-</td> <td>-</td> <td>326</td> <td>387</td> <td>303</td> </tr> </tbody> </table> <p> </p> <p>For modeling, we recommend using a train-test split where the test data includes geospatially distinct areas, thereby ensuring the model's ability to generalize to unseen regions is tested. We propose five cities varying in population size and from different regions in Germany for testing - images are tagged accordingly.</p> <p>Number of test images (train-test split): 776</p> <h3>Inter-rater-reliablility</h3> <p>Three annotators labeled the dataset, such that each image was annotated by one person. Annotators were encouraged to consult each other for a second opinion when uncertain.<br>1,800 images were annotated by all three annotators, resulting in a <em>Krippendorff's alpha</em> of 0.96 for surface type and 0.74 for surface quality.</p> <h3>Recommended image preprocessing</h3> <p>As the focal road located in the bottom center of the street-level image is labeled, it is recommended to crop images to their lower and middle half prior using for classification tasks.</p> <p>This is an exemplary code for recommended image preprocessing in <strong>Python</strong>:</p> <pre><code>from PIL import Image<br></code><code>img = Image.open(image_path)</code><br><code>width, height = img.size</code><br><code>img_cropped = img.crop((0.25 * width, 0.5 * height, 0.75 * width, height))</code></pre> <h3><br><strong>License</strong></h3> <p><a href="https://creativecommons.org/licenses/by-sa/4.0/">CC-BY-SA</a></p> <p> </p> <h3><strong>Citation</strong></h3> <p>If you use this dataset, please cite as: </p> <p> </p> <p>Kapp, A., Hoffmann, E., Weigmann, E. <em>et al.</em> StreetSurfaceVis: a dataset of crowdsourced street-level imagery annotated by road surface type and quality. <em>Sci Data</em> <strong>12</strong>, 92 (2025). https://doi.org/10.1038/s41597-024-04295-9</p> <p> </p> <p><code>@article{kapp_streetsurfacevis_2025,<br> title = {{StreetSurfaceVis}: a dataset of crowdsourced street-level imagery annotated by road surface type and quality},<br> volume = {12},<br> issn = {2052-4463},<br> url = {https://doi.org/10.1038/s41597-024-04295-9},<br> doi = {10.1038/s41597-024-04295-9},<br> pages = {92},<br> number = {1},<br> journaltitle = {Scientific Data},<br> shortjournal = {Scientific Data},<br> author = {Kapp, Alexandra and Hoffmann, Edith and Weigmann, Esther and Mihaljević, Helena},<br> date = {2025-01-16},<br>}</code></p> <p> </p> <p>-----------------------------------------------------------------------------------------------------------------------------------------------------------</p> <p>This is part of the SurfaceAI project at the University of Applied Sciences, HTW Berlin.</p> <p><br>- Prof. Dr. Helena Mihajlević<br>- Alexandra Kapp<br>- Edith Hoffmann<br>- Esther Weigmann</p> <p>Contact: surface-ai@htw-berlin.de</p> <p>https://surfaceai.github.io/surfaceai/</p> <p><strong>Funding</strong>: SurfaceAI is a mFund project funded by the Federal Ministry for Digital and Transportation Germany.</p> <p> </p>
Quantitative Content Analysis Data for Hand Labeling Road Surface Conditions in New York State Department of Transportation Camera Images
<p><strong>Foundational Codebook and Data: </strong></p> <p>Traffic camera images from the New York State Department of Transportation (511ny.org) are used to create a hand-labeled dataset of images classified into to one of six road surface conditions: 1) severe snow, 2) snow, 3) wet, 4) dry, 5) poor visibility, or 6) obstructed. Six labelers (authors Sutter, Wirz, Przybylo, Cains, Radford, and Evans) went through a series of four labeling trials where reliability across all six labelers were assessed using the Krippendorff’s alpha (KA) metric (Krippendorff, 2007). The online tool by Dr. Freelon (Freelon, 2013; Freelon, 2010) was used to calculate reliability metrics after each trial, and the group achieved inter-coder reliability with KA of 0.888 on the 4th trial. This process is known as quantitative content analysis, and three pieces of data used in this process are shared, including: 1) a PDF of the codebook which serves as a set of rules for labeling images, 2) images from each of the four labeling trials, including the use of New York State Mesonet weather observation data (Brotzge et al., 2020), and 3) an Excel spreadsheet including the calculated inter-coder reliability (ICR) metrics and other summaries used to asses reliability after each trial. The data are included in NYSDOT_quantitative_content_analysis.zip.</p> <p>The broader purpose of this work is that the six human labelers, after achieving inter-coder reliability, can then label large sets of images independently, each contributing to the creation of larger labeled dataset used for training supervised machine learning models to predict road surface conditions from camera images. The xCITE lab (xCITE, 2023) is used to store camera images from 511ny.org, and the lab provides computing resources for training machine learning models.</p> <p><strong>Obstructed Class Variation: </strong></p> <p>There are many applications for labeling roadside camera images, and as a variation of the foundational codebook, an addendum codebook provides another version of labeling the obstructed class. Specifically, this variation prioritizes labeling an image as “obstructed” only in extreme circumstances where there is a camera- or image- specific problem that prevents the assessment of any road surfaces. For labelers who want to use this version of the obstructed class (in this document) and also the other five weather-related classes (in the foundational codebook), the guidance is to use both documents in tandem, making sure to use the obstructed rules/definitions in this document while disregarding the obstructed rules/definitions in the foundational codebook. Alternatively, this codebook may be used alone in applications where the goal is to solely classify obstructed vs not obstructed. To ensure reliability and quality of this variation, quantitative content analysis was conducted on this addendum codebook, just as it was for the foundational codebook. Two labelers were tested with a sample of 30 images and achieved inter-coder reliability with Krippendorff's Alpha of 0.934 after one trial. The data, including the addendum codebook and labeling trial data (images and results) are included in ObstructedVariation_quantitative_content_analysis.zip.</p> <p>This material is based upon work supported by the U.S. National Science Foundation under Grant No. RISE-2019758.</p>
Source data for road transportation applications (road surface assessment, authentication of automotive vehicles)
<p>This data set records the driving using an Inertial Measurement Units of 12 different vehicles on the road infrastructure of the European Commission Joint Research Centre.</p> <p>The data set is described more in detail in the paper:</p> <p>Baldini, G.; Geib, F.; Giuliani, R. Continuous Authentication of Automotive Vehicles Using Inertial Measurement Units. <em>Sensors</em> <strong>2019</strong>, <em>19</em>, 5283.</p> <p><a href="https://doi.org/10.3390/s19235283">https://doi.org/10.3390/s19235283</a></p> <p>Please, cite this paper if you use this data set.</p>
SROADEX: Dataset for binary recognition and semantic segmentation of road surface areas from high resolution Aerial Orthoimages Covering Approximately 8,650 km2 of the Spanish Territory Tagged with Road Information
<p>The data have been generated using scripts developed in Python using Open Source libraries (GDAL/OGR and MapScript) for rasterization of vector cartography representing the axes of the different types of roads (urban, interurban and rural). This cartography has been obtained from different Spanish official sources (National Geographic Institute and autonomic cartographic agencies) that we have revised and edited in a meticulous and systematic way to verify that the roads are represented on the cartography according to the orthoimages, available on January 1, 2021 in the download center of the National Center of Geographic Information (CNIG), on 16 rectangular areas (28,5 km * 18,5 km) of the Spanish territory (insular and peninsular).</p> <p>The dataset consists of 777599 images in png format of 256x256 pixels, organized in folders for the different trainings, separating those corresponding to training, testing and validation.</p> <p>The structure of the data is as follows:<br> 1-Road-Ortho and 1-Road-Mask contain the images and ground true for training the semantic segmentation networks.<br> 1-Road-Ortho and 2-NoRoad-Ortho contain aerial images containing or not containing vials, for the training of binary tessellation networks identifying tessellations with vials.<br> Moreover, in each folder the structure is the same: train, test, validation containing 90%, 5% and 5% of the total images and masks of each type.</p> <p>1-Road-Ortho</p> <p> |----Train</p> <p> |----Test</p> <p> -----Validation</p> <p>1-Road-Mask</p> <p> |----Train</p> <p> |----Test</p> <p> -----Validation</p> <p>2-NoRoad-Ortho</p> <p> |----Train</p> <p> |----Test</p> <p> -----Validation</p> <p> </p>
Train and Evaluation Code, Road Classification Models and Test set of the paper "Insights into the Effects of Image Overlap and Image Size on Semantic Segmentation Models Trained for Road Surface Area Extraction from Aerial Orthophotography"
<p>This repository contains the Python scripts built for training and evaluation of the implementation, together with the test data and the resulting road segmentation models corresponding to the paper "Insights into the Effects of Image Overlap and Image Size on Semantic Segmentation Models Trained for Road Surface Area Extraction from Aerial Orthophotography". The scripts make use of the Tensorflow with Keras framework and their additional required dependencies.</p> <p>The training and validation set is based on the binary SROADEX dataset (<a href="../records/6482346">https://zenodo.org/records/6482346</a>) that was re-split into tiles that feature the image resolutions (256 x 256, 512 x 512, and 1024 x 1024 pixels) and image overlaps (0% and 12.5%) considered in this study. The data have been generated using scripts developed in Python using Open Source libraries (GDAL/OGR and MapScript) for rasterization of vector cartography that represents the axes of the different types of roads (urban, interurban and rural). This binary road data contains information from 16 full orthoimages (28.5 km * 18.5 km) with spatial resolution of 0.5 m/pixel from the insular and peninsular Spanish territory. Due to the size on disk of approximately 492 gigabytes, this training and validation data is only available upon request from the corresponding author. The test set has been generated from a novel area from Palencia (Spain) and features 18 million pixels labelled with the positive "Road" class. The test sets are provided in the repository for each resolution (with no overlap), so that additional DL models can be evaluated on the same data and compared with the results achieved in this study.</p> <p>The structure of the information shared in this repository is as follows:<br>The scripts have been grouped by tile resolution (256, 512 and 1024). First, the test set and the evaluation script can be found. For each tile resolution, there are two subfolders (corresponding to the "no overlap" and "12.5% overlap"). In each case, the Python scripts for training the models in the three repetitions are shared, and the trained models (H5 format) are shared in compressed form. Finally, for each resolution we also share the testing dataset which consists of two folders.</p> <p>The material is distributed under a CC-BY 4.0 license.</p>
Road surface temperature forecast study HKI-TKU 0708
<p>This dataset includes road weather station measurements, radar and HARMONIE forecast data related<br> to manuscript entitled "Verification of road surface temperature forecasts utilizing data from mobile sensors".<br> The manuscript will be submitted to a scientific journal for publication. Road weather model output<br> data is also included.</p> <p>Each folder contains ReaMe file for the folder's data.</p>
SmartRoadSense: aggregated road surface roughness dataset
<p>Dataset of aggregated road surface quality data points, collected through the mobile crowdsensing application SmartRoadSense.</p>
Stanford DAS-2 Array: Sand Hill Road Near-Surface Imaging Demo Dataset
<p>An excerpt of recordings from the Stanford DAS-2 array along Sand Hill Road, acquired on Dec 23rd, 2022, using OptaSense ODH-3, is shared here. The data is used in the demo of the newly proposed "target" interferometry method with traffic-induced surface waves for near-surface characterization. The sampling rate is 250 samples per second. The repository for this new method can be found at <a href="https://github.com/syyuan93/vehicle_tracking_based_time_lapse_monitoring">https://github.com/syyuan93/vehicle_tracking_based_time_lapse_monitoring</a>.</p>
ScienceDex guides
Understand access before you commit
These curated guides explain access requirements, typical timelines, costs, and reuse considerations for widely used research datasets.
Allen Brain Atlas
Allen Brain Atlas is an Allen Institute collection of brain map atlases, datasets, APIs, and analysis tools covering mouse, human, and non-human primate brain resources.
Annotated Behaviour and Observability Dataset (ABODe)
ABODe is a University of Edinburgh DataShare dataset for behavior classification in group-housed mice using home-cage video, identities, bounding boxes, ground-plate positions, and annotator labels.
DANDI Archive for NWB datasets
DANDI is a BRAIN Initiative archive for publishing and sharing neurophysiology data, including electrophysiology, optophysiology, and behavioral data packaged as NWB and related standards.
International Brain Laboratory public data
The International Brain Laboratory public data releases expose standardized mouse decision-making experiments, including Neuropixels recordings, widefield calcium imaging, behavior, and session metadata accessed through the ONE API.
OpenNeuro
OpenNeuro is a free, open platform for sharing neuroimaging datasets, with public search, dataset pages, and download paths for web, S3, DataLad, and the OpenNeuro CLI.