Recognition: vehicles
Cluster cars, bikes, boats and horses; manage drivers.
Motorsport, cycling, sailing and equestrian events are about the machine as much as the person. Vehicle detection clusters cars, motorbikes, bikes, boats and horses by appearance, so you can pull every shot of one car — even across corners, laps and lighting — without reading a single number.
Turn it on
On the album's Edit → Workflows page, under People detection, switch on Vehicle detection. Beta Web SnapFlow groups visually similar vehicles as photos are ingested, using the whole-vehicle appearance (livery, shape, colour) rather than just a number plate.
Vehicles and their drivers
A car number on the numbers detail page can be stored as a vehicle, which persists it for cross-event tracking. A vehicle can have more than one driver — endurance line-ups, co-drivers, reserves — and the detail page's drivers panel lets you add or remove each one, optionally with a role (e.g. "co-driver"). Every photo of the vehicle then tags all of its drivers, so each driver's gallery is complete.
Vehicles are typed by kind — car, bike, boat or horse — which keeps the right terminology and matching behaviour for each sport. (Sailing boats and event horses are first-class here, not bolted onto "car".)
Correcting
As with the other pipelines, you stay in charge:
- Set / change the vehicle a photo belongs to.
- Ignore vehicles you don't want clustered (a safety car, a course vehicle).
- Manage the driver line-up from the vehicle's panel.
Pair it with numbers
Vehicle clustering and number reading reinforce each other: the number names the entry, the visual cluster catches the frames where the number is hidden. Together they give you every shot of car #55 — plate visible or not.
Next: identifying drivers and riders by their helmet.