DOCSEARCHbyAlgolia
Submit Docs

WebODM Documentation

website·Indexed 33 days ago

https://docs.webodm.net

Documentation for WebODM and WebODM Lightning, covering aerial photogrammetry workflows including image capture, task creation, and 3D reconstruction. Includes guides on using Ground Control Points (GCPs), processing thermal images and video files, calculating volumes, generating elevation profiles, and managing coordinate reference systems.

Tokens
14.7K
Snippets
16
Records
141
Agent score
58%

What's inside WebODM

  1. Overview of Gaussian Splatting in WebODM

    Gaussian splatting creates photorealistic 3D scenes using millions of semi-transparent 3D ellipses. In WebODM, splats are generated by training a model using images, camera poses, and sparse points exported from WebODM Lightning. Once trained, the resulting splat model can be uploaded to WebODM Lightning for display alongside point clouds and textured models.

    Critical Requirement: To achieve good results, datasets must include oblique or orbit shots. Nadir-only flights typically yield poor results.

  2. Crop map overviews for annotation reports

    To focus the cover page or overview images on a specific area instead of the entire map, use the Crop button to define a crop area before exporting the PDF.

    To undo a crop after exporting, click the Crop button again and then right-click.

  3. Exclude image areas using Image Masks

    Image masking allows you to specify portions of an image to be excluded during reconstruction to eliminate noise or undesired elements. Black areas in a mask image are ignored during the reconstruction process. To effectively remove an item, it must be masked in every single image where it appears.

  4. Understand the role of Ground Control Points (GCPs) in WebODM

    Ground Control Points (GCPs) are precise real-world position measurements (latitude/longitude) taken on the ground, typically using high-accuracy GNSS devices. They are critical for models where images are not georeferenced; without GCPs, such models may be improperly oriented or scaled. To achieve optimal accuracy, it is recommended to use 5 to 8 GCPs evenly distributed across the surveyed area. Each marker should ideally be captured in at least 3 images and tagged in every image where it is visible.

  5. Use Ground Control Points in WebODM

    Ground Control Points (GCPs) are used to improve the accuracy of reconstructions by uploading a specific text file named gcp_list.txt along with the input images when creating a task in WebODM Lightning.

  6. Understand and identify the Bowl Effect in photogrammetry

    The bowl effect (also known as the dome effect) is a common distortion in aerial photogrammetry where the reconstructed 3D model curves upward or downward at the edges rather than remaining flat. This is typically caused by suboptimal flight paths, excessive lens distortion, or lack of sufficient spatial constraints.

  7. Override or add GPS image data using a geo.txt file

    To provide missing GPS data or override existing GPS data for input images in WebODM, create a geolocation file named exactly geo.txt. This file should be uploaded along with the images. For data using latitude, longitude, and height above the WGS84 ellipsoid, the file must start with the coordinate system identifier EPSG:4326, followed by lines containing the image filename, longitude, latitude, and altitude.
    EPSG:4326
IMG_0001.JPG longitude latitude altitude
IMG_0002.JPG longitude latitude altitude

  8. Import GCP measurements via CSV

    To import GCP measurements into the GCP Interface, upload a CSV file containing the target locations. The software automatically recognizes columns based on the following keywords:

    • GCP Label: GCP Label, Label, Name, GCP
    • X (Easting): Easting, X, Longitude, Lon
    • Y (Northing): Northing, Y, Latitude, Lat
    • Z (Elevation): Elevation, Z, Altitude, Alt

    After selecting the CSV, you must provide the coordinate reference system (CRS) using an EPSG code or a PROJ.4 string definition.

    GCP Label,Northing,Easting,Elevation
GCP-1,235269.88,3811198.11,0.20
GCP-2,235246.25,3811220.60,0.25

  9. Automatically align two overlapping datasets in WebODM Lightning

    WebODM Lightning allows the automatic alignment of two overlapping datasets without requiring ground control points (GCPs). This feature is primarily used for comparing changes over time or compositing larger maps. To align datasets, both tasks must be located within the same project and must have overlapping imagery; otherwise, the alignment option will not be available in the interface.

  10. Use spatial constraints to mitigate the Bowl Effect

    Providing absolute reference points helps anchor the reconstruction and prevents compounding errors that lead to the bowl effect:

    • Ground Control Points (GCPs): Place surveyed markers with known coordinates throughout the scene to provide absolute reference points.
    • RTK/PPK Positioning: Use drones equipped with Real-Time Kinematic (RTK) or Post-Processed Kinematic (PPK) sensors to provide centimeter-accurate location data for each image, reducing the reliance on GCPs.

  11. Share a WebODM Map or 3D Model

    To share the results of a processed dataset from the WebODM cloud interface, navigate to the specific result type (Map or 3D Model) and use the Share button located in the bottom-right corner to generate a shareable link.

  12. Collect Ground Control Points via field measuring

    To collect GCPs in the field, place visible markers evenly across the survey area (avoiding edges). Use a GNSS device to record the coordinates of each marker. Ensure there is a clear line of sight between the camera and the marker, as obstructions like trees or buildings can degrade accuracy. It is recommended to use a CSV template from the GCP Interface for recording these measurements.