Import video frames
The Import video frames dialog allows you to extract frames from a video file and add them to the block.
To extract frames from a video, enter the input video file and select the import settings:
- Start time/end time: allows you to import only a section of the input video (by default all the video will be imported).
- Extract a photo every: sampling interval, in seconds. This defines the number of photos that will be extracted from the input video sequence.
- Photo output directory: directory where the extracted photos are created.
Please select the video sampling interval to ensure a correct overlap between photos.
Focal length variations are not supported.
All imported frames are added in one unique photogroup; ContextCapture assumes that the same camera model (with the same optical properties) is used during all the imported sequence. If the zoom varies during the video sequence, please use the start/end settings in our import to split the video in sections of constant zoom and to import each section separately.
Click on Import to extract and add the video frames.
Once imported, extracted photos can be inspected individually from the block's Photos page.
Frames imported from a video files have an unknown camera model. We recommend you to define the photogroup's main optical properties (sensor size, focal length and camera model type) before starting an aerotriangulation. One option for doing this is to search for a suitable camera model in the camera database (Camera Database). See also Photos.
In case no camera model in the database fits your camera, you can find your camera model attributes by using the self-calibration property of ContextCapture 's aerotriangulation. For optimal results, we recommend you to use a 360° video of a highly textured and geometrically complex object. The camera parameters obtained after self-calibration can be stored in the camera database and reused for your more complex projects.