Capability 03 · Vision
Turn camera footage into records you can search.
Our vision models find and follow the people and vehicles on camera, and turn hours of footage into records your team can search and count.
Live record
| Track | Class | X | Y | Conf |
|---|---|---|---|---|
Each row here comes from a box you can see on the footage. Positions are in source pixels, at the bottom centre of each box.
What goes in
- Fixed cameras and existing CCTV
- Recorded footage and archives
- Drone, vehicle and body-worn video
- Still imagery and scans
What comes out
- People, vehicles and objects detected and followed frame to frame
- Positions and paths over time
- Events with a timestamp and a confidence
- Every record tied to the frame it came from
Put it to work on your data.
Tell us what you have and the decision it should support. We'll show you what this looks like on it.
Talk to us