Photogrammetry Integration

DroneDB Registry integrates with photogrammetry processing platforms (NodeODX, OpenDroneMap, ODX) to send drone images for processing and retrieve the results directly into a dataset.

How it works

The Processing Platform in Registry can submit a batch of images to a remote NodeODX/ODX processing node. The node processes the images and produces outputs (orthophoto, DEM, point cloud, 3D model, etc.). The results are downloaded as a ZIP archive (typically all.zip from NodeODX) and extracted directly into the source dataset, or into a newly created dataset. Extracted files are indexed and per-file build jobs are enqueued automatically.

Processing nodes

A processing node is a remote instance of NodeODX, OpenDroneMap, or ODX that accepts processing jobs via API. Nodes are registered in Registry with:

Setting	Description
Node ID	Unique identifier for the node
URL	Base URL of the processing API
API key	Authentication token for the node
Name	Human-readable label

In Hub

Submit a photogrammetry job

1. Open a dataset containing drone images
2. Select the images to process (or use all images in the dataset)
3. Click the Photogrammetry toolbar button or processing task
4. Select a processing node and configure options
5. Submit the job

The job runs as a background Hangfire task. Progress and logs are streamed to the task panel.

Processing options

The Registry photogrammetry tool accepts the following parameters:

Parameter	Type	Description
folder	string	Dataset folder to scan for images (default: whole dataset)
images	string[]	Explicit list of image entry paths to process
nodeId	string	Target NodeODX node id (default: first configured node)
name	string	Task name shown on the node
options	{name, value}[]	NodeODX/ODX options array passed verbatim to the processing node (e.g. [{"name":"dsm","value":true}])
destPath	string	Destination folder within the dataset for extracted results (required unless createNewDataset is true)
createNewDataset	boolean	If true, create a new dataset for the results instead of extracting into the current dataset
newDatasetName	string	Slug for the new dataset (kebab-case, max 128 chars; auto-generated if omitted)
newDatasetOrgSlug	string	Organization slug for the new dataset (default: same as source)
newDatasetVisibility	string	Visibility for the new dataset: PRIVATE, UNLISTED, or PUBLIC (default: PRIVATE)

Common NodeODX options to place inside the options array:

Option	Example	Description
dsm	{"name":"dsm","value":true}	Generate Digital Surface Model
dtm	{"name":"dtm","value":true}	Generate Digital Terrain Model
orthophoto	{"name":"orthophoto","value":true}	Generate orthophoto
pc	{"name":"pc","value":true}	Generate point cloud
3d-tiles	{"name":"3d-tiles","value":true}	Generate 3D tiles
feature-quality	{"name":"feature-quality","value":"high"}	Processing quality (low, medium, high)
gsd	{"name":"gsd","value":0.01}	Estimated ground sample distance

Task output

When processing completes, the result bundle is extracted directly into the dataset (or into the newly created dataset). No separate downloadable artifact is produced. Extracted files are indexed automatically and their derivative builds (COG, COPC, NXS, 3D Tiles, etc.) are enqueued in the background.

Workflow example

1. Upload drone images to a Hub dataset
2. Submit photogrammetry job via processing platform
3. Wait for processing to complete (monitor via task panel)
4. Results are extracted and indexed automatically in the destination dataset
5. View orthophoto, DEM, point cloud, and 3D model in Hub

tip

For large datasets, consider using a dedicated processing node with sufficient RAM and GPU resources. The processing platform supports multiple nodes for load balancing.

CLI alternative

For offline photogrammetry processing, use WebODM or ODX directly and then push the results to DroneDB:

# Process with ODX/WebODM locally
# Then add results to DroneDB
ddb add odm_orthophoto/ odm_dem.tif odm_3d_model/

# Push to Hub
ddb push