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.
222
datasets available to search
ShareScore release 0.7.1
Dataset results
222 results for “point cloud”
Fatiando a Terra Data: British Columbia, Canada - Lidar point cloud
<p>This is a point cloud sliced to the small <a href="https://apps.gov.bc.ca/pub/bcgnws/names/21973.html">Trail Islands</a> to the North of Vancouver to reduce the data size. The islands have some nice looking topography and their isolated nature creates problems for some interpolation methods, like simple nearest neighbours.</p> <p><strong>Note:</strong> This is a processed and formatted version of the source dataset below. It's meant for use in documentation and tutorials of the <a href="https://www.fatiando.org">Fatiando a Terra</a> project. Please <strong>cite the original authors</strong> when using this dataset.</p> <p><strong>Changes made: </strong>Cut out a section of the original data to focus only on these small islands; Change UTM coordinates to geodetic latitude and longitude (WGS84); Keep only the ground reflection picks; Export to a compressed CSV for easier loading with Pandas.</p> <p><strong>Source: </strong><a href="https://www2.gov.bc.ca/gov/content/data/geographic-data-services/lidarbc">LidarBC</a></p> <p><strong>Source license: </strong><a href="https://www2.gov.bc.ca/gov/content/data/open-data/open-government-licence-bc">Open Government Licence - British Columbia</a></p> <p><strong>Repository:</strong> <a href="https://github.com/fatiando-data/british-columbia-lidar">https://github.com/fatiando-data/british-columbia-lidar</a></p>
Terrestrial and aerial photos, GCPs and derived point clouds of a sinkhole in Northern Thuringia
<p>Aiming at comparing the results of using either terrestrial or aerial photos for structure from motion photogrammetry of a sinkhole in Northern Thuringia we took these photos in 2017, 2018 and 2019 and surveyed ground control points. Therefore, the data allows both, the comparison of the results of using terrestrial or aerial photos for the 3D reconstruction, and the mulit-year monitoring of geomorphic changes within the sinkhole.</p>
3D-MSNet: A point cloud based deep learning model for untargeted feature detection and quantification in profile LC-HRMS data
<p>Supplementary data of 3D-MSNet</p>
Figure 4 from: Kissling WD, Seijmonsbergen AC, Foppen RPB, Bouten W (2017) eEcoLiDAR, eScience infrastructure for ecological applications of LiDAR point clouds: reconstructing the 3D ecosystem structure for animals at regional to continental scales. Research Ideas and Outcomes 3: e14939. https://doi.org/10.3897/rio.3.e14939
Figure 4 - Time table for the eEcoLiDAR project (assuming a start in March 2017). The work plan covers tasks for the NLeSC engineers, the proposed PhD student, and two associated Postdoc projects.
Figure 3 from: Kissling WD, Seijmonsbergen AC, Foppen RPB, Bouten W (2017) eEcoLiDAR, eScience infrastructure for ecological applications of LiDAR point clouds: reconstructing the 3D ecosystem structure for animals at regional to continental scales. Research Ideas and Outcomes 3: e14939. https://doi.org/10.3897/rio.3.e14939
Figure 3 - Example of identifying trees in a forest from LiDAR data. Illustrated is a small plot of poplar trees in Flevoland, The Netherlands, for which tree crowns and tree tops have been calculated.
Figure 2 from: Kissling WD, Seijmonsbergen AC, Foppen RPB, Bouten W (2017) eEcoLiDAR, eScience infrastructure for ecological applications of LiDAR point clouds: reconstructing the 3D ecosystem structure for animals at regional to continental scales. Research Ideas and Outcomes 3: e14939. https://doi.org/10.3897/rio.3.e14939
Figure 2 - Generic workflow for object-based image analysis (OBIA) of LiDAR point clouds and proposed ecological applications. A workbench (blue) will be developed to handle the data storage, data exploration, and interactive OBIA of the massive LiDAR point clouds. Combined with datasets of bird distributions, climate, and other remote sensing layers (orange), the LiDAR data will be applied to several ecological case studies, e.g. by using species distribution modelling of birds and insect pollinators (green).
Figure 1 from: Kissling WD, Seijmonsbergen AC, Foppen RPB, Bouten W (2017) eEcoLiDAR, eScience infrastructure for ecological applications of LiDAR point clouds: reconstructing the 3D ecosystem structure for animals at regional to continental scales. Research Ideas and Outcomes 3: e14939. https://doi.org/10.3897/rio.3.e14939
Figure 1 - The vertical and horizontal distribution of plants influences habitat structure and 3D characteristics of vegetation for animals. Illustrated are examples for (a) forests, (b) agricultural and open landscapes, and (c) reedbeds and marshlands. The height, openness and density of vegetation as well as specific habitat features (e.g. tree species, hedges etc.) are key aspects of animal habitat and space use.
Senator Beck Basin Lidar Point Cloud 2017-02-21
<p>This lidar dataset was collected using an airborne Riegl Q1560 dual laser scanner flown aboard a King Air A90 as part of the Airborne Snow Observatory (ASO), a now complete NASA-JPL project. The data acquisition took place on February 21, 2017, over the Senator Beck Basin Study Area in the San Juan Mountains, Colorado as part of the NASA SnowEx campaign (Year 1). Ackroyd et al. (under review) utilized this point cloud for a lidar intensity correction to retrieve snow cover extent and grain size in GIScience and Remote Sensing.</p>
InLUT3D: Indoor Lodz University of Technology Point Cloud Dataset
<h2>Background</h2> <p>This resource contains Indoor Lodz University of Technology Point Cloud Dataset (<strong>InLUT3D</strong>) - a point cloud dataset tailored for real object classification and both semantic and instance segmentation tasks. Comprising of <strong>321</strong> scans, some areas in the dataset are covered by multiple scans. All of them are captured using the Leica BLK360 scanner.</p> <h2>Train/test split</h2> <p>The datset's authors impose the following train-test split:</p> <table> <tbody> <tr> <td><strong>Split</strong></td> <td><strong>Setup range<br></strong></td> </tr> <tr> <td>train</td> <td>from <em>setup_0 </em>to <em>setup_300</em></td> </tr> <tr> <td>test</td> <td>from <em>setup_301 </em>to <em>setup_320</em></td> </tr> </tbody> </table> <p>The corresponding setups' objects are recommended for splits for the classificcation task.</p> <h2>Available categories</h2> <p>The points are divided into <strong>18 </strong>distinct categories outlined in the <em>label.yaml</em> file along with their respective codes and colors. Among categories you will find:</p> <ul> <li>ceiling,</li> <li>floor,</li> <li>wall,</li> <li>stairs,</li> <li>column,</li> <li>chair,</li> <li>sofa,</li> <li>table,</li> <li>storage,</li> <li>door,</li> <li>window,</li> <li>plant,</li> <li>dish,</li> <li>wallmounted,</li> <li>device,</li> <li>radiator,</li> <li>lighting,</li> <li>other.</li> </ul> <h2>Challenges</h2> <p>Several challenges are intrinsic to the presented dataset:</p> <ol> <li>Extremely non-uniform categories distribution across the dataset.</li> <li>Presence of virtual images, particularly in reflective surfaces, and data exterior to windows and doors.</li> <li>Occurrence of missing data due to scanning shadows (certain areas were inaccessible to the scanner's laser beam).</li> <li>High point density throughout the dataset.</li> </ol> <h2>Dataset structure</h2> <p>The structure of the dataset is the following:</p> <p>inlut3d.tar.gz/<br>├─ setup_0/<br>│ ├─ projection.jpg<br>│ ├─ segmentation.jpg<br>│ ├─ setup_0.pts<br>├─ setup_1/<br>│ ├─ projection.jpg<br>│ ├─ segmentation.jpg<br>│ ├─ setup_1.pts<br>...</p> <table> <tbody> <tr> <td><strong>projection.jpg</strong></td> <td>A file containing a spherical projection of a corresponding PTS file.</td> </tr> <tr> <td><strong>segmentation.jpg</strong></td> <td>A file with objects marked with unique colours.</td> </tr> <tr> <td><strong>setup_x.pts</strong></td> <td>A file with point cloud int the textual PTS format.</td> </tr> </tbody> </table> <h2>Point characteristic</h2> <p>Each PTS file contains 8 columns:</p> <table> <tbody> <tr> <td><strong>Column ID</strong></td> <td><strong>Description</strong></td> </tr> <tr> <td>1</td> <td>X Cartesian coordinate</td> </tr> <tr> <td>2</td> <td>Y Cartesian coordinate</td> </tr> <tr> <td>3</td> <td>Z Cartesian Coordinate</td> </tr> <tr> <td>4</td> <td>Red colour in RGB space in the range [0, 255]</td> </tr> <tr> <td>5</td> <td>Green colour in RGB space in the range [0, 255]</td> </tr> <tr> <td>6</td> <td>Blue colour in RGB space in the range [0, 255]</td> </tr> <tr> <td>7</td> <td>Category code</td> </tr> <tr> <td>8</td> <td>Instance ID</td> </tr> </tbody> </table> <p> </p>
NIBIO_MLS: a forest point cloud panoptic segmentation dataset from mobile laser scanning (Geoslam Horizon)
<h1>General description</h1> <p>This dataset consists of a ML-ready labelled mobile laser scanning (MLS) point cloud dataset including 16 manually labelled forest plots (approx. 250 m2) for forest panoptic segmentation and thus including both semantic and instance labels. The data was collected using a Geoslam Horizon RT and processed using Geoslam Hub.</p> <h1>Labels</h1> <p>The data were then labelled into the following semantic classes (<em>label</em>):</p> <ul> <li>1= ground</li> <li>2= vegetation: these include both branches, leaves, and low vegetation</li> <li>3= lying deadwood</li> <li>4= stems</li> </ul> <p>In addition for each tree, a unique tree identifier (<em>treeID</em>) was also assigned to each point.</p> <h1>Data split</h1> <p>Each plot was split into train (50%), validation (25%), and test (25%) sets by dividing the circular plot into four slices, out of which the first two were used for training, the third for validation, and the fourth for test. </p> <p>Thus the users might play around with merging the train and validation dataset as they prefer. These two sets can be used during model training, hyperparameter tuning, and model selection. However, the test set should be kept as an independent set to be used for benchmarking against the values reported in the two studies indicated below. </p> <h1>Citation</h1> <p>To cite this datasets and for a more detailed description use:</p> <p>Wielgosz, M., Puliti, S., Xiang, B., Schindler, K. and Astrup, R., 2024. SegmentAnyTree: A sensor and platform agnostic deep learning model for tree segmentation using laser scanning data. <em>Remote Sensing of Environment; </em></p> <h2>Other studies using these data</h2> <p>Wielgosz, M., Puliti, S., Wilkes, P. and Astrup, R., 2023. Point2Tree (P2T)—Framework for parameter tuning of semantic and instance segmentation used with mobile laser scanning data in coniferous forest. <em>Remote Sensing</em>, <em>15</em>(15), p.3737; available <a href="https://www.mdpi.com/2072-4292/15/15/3737" target="_blank" rel="noopener">here</a></p> <h1>Funding</h1> <p>This work is part of the Center for Research-based Innovation SmartForest: Bringing Industry 4.0 to<br>the Norwegian forest sector (NFR SFI project no. 309671, smartforest.no).</p> <h1>⚖️ Licensing</h1> <p>📄 Please refer to the specific licenses below for details on how the data can be used.</p> <h4>🔑 Key Licensing Principles:</h4> <ul> <li>✅ You may access, use, and share the dataset and models freely.</li> <li>🔄 Any derivative works (e.g., trained models, code for training, or prediction tools) must also be made publicly available under the same licensing terms.</li> <li>🌍 These licenses promote <strong>collaboration</strong> and <strong>transparency</strong>, ensuring that research using this dataset benefits the broader scientific and open-source community 🙌</li> </ul>
MuSHRoom iPhone dataset colmap pose & point cloud
Open the record for dataset details and reuse information.
RibSeg Dataset and Strong Point Cloud Baselines for Rib Segmentation from CT Scans
<p>Manual rib inspections in computed tomography (CT) scans are clinically critical but labor-intensive, as 24 ribs are typically elongated and oblique in 3D volumes. Automatic rib segmentation methods can speed up the process through rib measurement and visualization. However, prior arts mostly use in-house labeled datasets that are publicly unavailable and work on dense 3D volumes that are computationally inefficient. To address these issues, we develop a labeled rib segmentation benchmark, named RibSeg, including 490 CT scans (11,719 individual ribs) from a public dataset. For ground truth generation, we used existing morphology-based algorithms and manually refined its results. Then, considering the sparsity of ribs in 3D volumes, we thresholded and sampled sparse voxels from the input and designed a point cloud-based baseline method for rib segmentation. The proposed method achieves state-of-the-art segmentation performance (Dice<span class="math-tex">\(\approx95\%\)</span>) with significant efficiency (<span class="math-tex">\(10\sim40\times\)</span> faster than prior arts). The RibSeg dataset, code, and model in PyTorch are available at <a href="https://github.com/M3DV/RibSeg">https://github.com/M3DV/RibSeg</a>.</p> <p> </p> <p><strong>Note:</strong> This repository provides rib segmentation ("RibFrac31-rib-seg.nii.gz") and centerline ("RibFrac31-rib-cl.nii.gz") <em>annotations</em> for 490 cases in RibFrac dataset. Please download the corresponding CT <em>images </em>("RibFrac31-image.nii.gz") at <a href="https://ribfrac.grand-challenge.org/">https://ribfrac.grand-challenge.org/</a> (1-click registration is needed via <em>"Join"</em>).</p>
Coit Tower - Point Cloud
Point cloud obtained from drone video Source: Objaverse 1.0 / Sketchfab
Data from: Object recognition and localization from 3D point clouds by maximum-likelihood estimation
Open the record for dataset details and reuse information.
S-MODE MASS Level 1 Lidar Point Cloud Version 1
This dataset contains geolocated airborne LiDAR point cloud measurements from the Sub-Mesoscale Ocean Dynamics Experiment (S-MODE) conducted approximately 300 km offshore of San Francisco during a pilot campaign over two weeks in October 2021, and two intensive operating periods (IOPs) in Fall 2022 and Spring 2023. S-MODE aims to understand how ocean dynamics acting on short spatial scales influence the vertical exchange of physical and biological variables in the ocean. The Modular Aerial Sensing System (MASS) is an airborne instrument package that is mounted on the DHC-6 Twin Otter aircraft which flies long duration detailed surveys of the field domain during deployments. MASS includes a high resolution LiDAR, used to characterize the properties of ocean surface topography. The sensor has a maximum pulse repetition rate of 400 kHz, with a +/- 30° cross-heading raster scan rate of 200 Hz. Level 1 LiDAR point clouds are available in .laz format.
G-LiHT Lidar Point Cloud V001
Goddard’s LiDAR, Hyperspectral, and Thermal Imager ([G-LiHT](https://gliht.gsfc.nasa.gov/)) mission is a portable, airborne imaging system that aims to simultaneously map the composition, structure, and function of terrestrial ecosystems. G-LiHT primarily focuses on a broad diversity of forest communities and ecoregions in North America, mapping aerial swaths over the Conterminous United States (CONUS), Alaska, Puerto Rico, and Mexico.The purpose of G-LiHT’s LiDAR Point Cloud data product (GLLIDARPC) is to provide high-density individual LiDAR return data, including 3D coordinates, classified ground returns, Above Ground Level (AGL) heights, and LiDAR apparent reflectance. GLLIDARPC data are processed as a LAS Version 1.1 binary format specified by the American Society for Photogrammetry and Remote Sensing (ASPRS). The point cloud includes a density of more than 10 points per square meter. A low resolution browse is also provided showing the LiDAR Point Cloud as an Inverse Data Weighted (IDW) interpolation in PNG format.
Separation, Preconcentration, and Determination of Lead, Cadmium and Iron Using Cloud Point Extraction with FAAS
<p>The concentration of heavy metals in drinking water is a significant standard for water quality evaluation and water pipeline corrosion detection. This research aimed to develop a new method based on cloud point extraction (CPE) for the separation, preconcentration and detection of lead, cadmium, and iron with flame atomic absorption spectrometry (FAAS) when the concentration of metallic trace elements in the sample was lower than the limits of detection (LOD). The experimental LODs for Pb, Cd and Fe, determined based on significant sensitivity change and break in the slope of the standard curve, were 0.01, 0.01, and 0.3 ppm, respectively. With joint utilization of both 2,6-diamino-4-phenyl-1,3,5-triazine (DPT) and 3-amino-7-dimethylamino-2-methylphenazine (Neutral Red, NR) as chelating agents, and using Triton X-114 as surfactant, those metallic elements could be effectively enriched in water samples. The preconcentration procedure was optimized by varying the experimental factors such as temperature, equilibrium time, pH, and the concentration of the chelates and surfactant. With the optimization, this method allowed the determination of these three trace elements with a 20-times reduced LOD and yielded substantial recoveries of 99.8, 97.3 and 99.3% for Pb, Cd and Fe, respectively.</p>
Data for EEG Electrode Localization with 3D iPhone Scanning Using Point-Cloud Electrode Selection (PC-ES)
<p>The dataset contains 1) the deidentified, cropped Heges scans collected on 8 subjects (4-7 sub-scans per subject), 2) the merged scans via the PC-ES software from 3 reviewers, and 3) the photogrammetry data (validation data) which also constructed from 3 reviewers. The data is in MATLAB data files (mat-files), which is consistent with the PC-ES software (https://github.com/haley-richards/PC-ES). There are also 2 m-files (MATLAB scripts) that can be used to plot the individual or merged scans. </p><ul><li>cropped_deidentified_scans_pg.mat<ul><li>cropdeidscans a 1 x 8 structure array (8 subjects)</li><li>Each cell contains a structure array of all the sub-scans for the given subject with the following fields:<ul><li>colobj – Matlab point cloud object. This is the Heges output cropped and deidentitied.</li><li>tri – triangulation related to the nodes (Location) of the colobj object</li></ul></li></ul></li><li>merged_scans – output from the PCES software that merged the subscans from Heges<ul><li>outputres a 3 x 8 structure array (3 reviewers/readers of scans and 8 subjects)</li><li>Each cell contains a structure array scans with the following fields:<ul><li>colobj – the merged scans, albeit separated by angular sectors</li><li>tdat_mrg – merged electrodes</li><li>tdat_full – full-set of electrodes</li><li>tri_mrg - triangulation related to the nodes (Location) of the colobj object</li></ul></li></ul></li><li>photogram_dat – photogrammetry-base electrodes 3 x 8 structure array (3 reviewers/readers of scans and 8 subjects)</li></ul>
Dataset of point-clouds generated by UAV LiDAR scanning of barley field-plot trial.
<p>Point clouds from eight scanning flights, scanned by UAV LiDAR system VUX-UAV1 (Ricopter GmBh), provided in forms of separated .las files. Dataset represent raw data that were further processed by software ALFA to extract height of canopies in particular field plots. Software serves for monitoring of plant growth dynamics in agricultural or breeding filed-plot trials.</p>
Scottish Boatbuilding School: point cloud (I)
The Scottish Boatbuilding School at the Scottish Maritime Museum in Irvine was established in 2014 to provide education and qualification in both traditional and modern boat building, as well as vessel conservation and restoration. This pointcloud generated from photogrammetry as a part of 'Scanning the Horizon' project, provides a unique glimpse into the school's boatbuilding workshop, which is not accessible for the public. Listen to Martin, the Boatbuilding School Manager as he talks about current restoration work happening at the workshop and unique tools of the traditional boatbuilding trade. To discover more about the boatbuilding workshop at the Scottish Maritime Museum visit: https://skfb.ly/6MyMF Source: Objaverse 1.0 / Sketchfab
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.