Extract Features is using detectors that are trained with machine learning, to automatically extract information from your photos, point clouds and meshes.
You can start extracting features from data that is part of your current ContextCapture project (e.g. photos, block, point cloud productions, reality meshes).
Alternatively, you can also Extract Features on reality data that are not within your ContextCapture project. These are more advanced workflows and require learning about ContextScene format.

Here is a dedicated video tutorial describing crack detection in 3D-world

Feature extraction workflow
The workflow for feature extraction consists in a few steps. In this example, 2D object detection is used but the principle is the same for all Feature extraction types.

1. Start the feature extraction workflow using the “plus” icon in your project tree view and click “Extract Features”

2. Select your feature extraction type.

3. Select the input data, the detector to be used and additional settings when required.

4. Press "Create”, this will add a new “analysis item” in the project tree view

5. Go to the settings tab of the Analysis item and adjust settings if needed (e.g. modify region of interest to detect only on a subpart of your data).

6. Press the Submit button to initiate the analysis job. Select your output format(s) and press Submit.

7. Review the progress in the Overview tab

8. Go to the “Result” tab to visualize your results when processing is complete. Results will show in 2D or 3D views depending of the type of outputs generated.