There are not a lot of video surveillance companies that survive 9 years with only one feature that makes their product stand out.
In the case of Snap Surveillance, the feature is their product, Snap Force Multiplier (FMX) which allows security operators to manually track subjects across a camera system within a single video tile.
In this note, based on a discussion with their CEO, we will be examining Snap's approach to market, platform capabilities, and challenges.
The system does not track any objects. It could learn the relationships between outdoor cameras and allow you to follow a vehicle across multiple cameras. You can also manually create links between cameras as well.
Can you clarify exactly what type of operation you were thinking about? There would be better solutions if you need automatic tracking.
Sean is correct. Snap is not an automated tracker, we keep "human in the loop".
The solution focuses on helping - guiding - operators to more quickly and reliably follow a target (person, vehicle or other object) from camera to camera through a large surveillance network. Snap learns camera network links, scaling up to 1,000+ cameras in a group, and uses this simplify the life of an operator when they need to either track a target live, or to review large sets of archived video (with almost no ongoing computing overhead).
We can be used in camera networks to help follow cars, but with a few qualifiers if its for traffic cameras or the like:
- traffic camera networks are often just on main street corners with a considerable distance between each and large gaps in coverage - that's not playing to Snap's strengths; and
- traffic cameras are often PTZs that get regularly moved around - Snap is of most benefit with a network of primarily stationary / fixed cameras