The technically talented and customer-focused crew at PhotoModeler has been making big strides in helping the recon community acquire accurate results, as easily as possible.

As discussed in this edition of To the Point, they developed a tool to quickly import vehicle point clouds and align them to corresponding pixels. A valuable workflow for determining speed from video, or final rest positions.

This summer, they made another major advancement: we can import videos directly into the program. This enables us to solve for camera parameters (location, focal length, distortion) and vehicle positions, while seamlessly generating a speed report within PhotoModeler.

Aside from performing a forensic evaluation/enhancement of the video (metadata, frame timing, macroblock analysis, deinterlacing, adjusting/editing, etc.), PhotoModeler provides a one-stop-shop for speed determination, whether the camera is stationary or moving.

Here’s the 30,000-foot view of the process for a stationary camera:

  1. Import video into PhotoModeler.

  2. Establish camera parameters using control geometry (3D coordinates of landmarks visible in the video).

  3. Navigate to the frames of interest and align the vehicle point cloud with corresponding pixels. This will automatically create a bookmark, detailing frame number, timing, and vehicle position.

  4. Export speed and acceleration report from PhotoModeler. Boom!

If you want to alter the frame timing reported in PhotoModeler, that flexibility exists. Hats off to the PhotoModeler team. The user interface and process are excellent. Check out the video below (or click here) to see the tool in action.

If you’d like to learn more about photogrammetry and determining speed from video using PhotoModeler and PFTrack, let me know. I expect to offer one more four-day photogrammetry course before the year’s end.

Thanks for reading, keep exploring!

Lou Peck
Lightpoint | JS Forensics