Mark,
PureTech Systems (puretechsystems.com) provides video analytics for tunnels, most recently the Tom Lantos tunnel for Caltrans in California. You can read details on that install here https://goo.gl/sFue2M
Here are some insights based on our experience:
1. Has anyone used video analytics to help monitor traffic conditions, and how well did it work?
The system was installed for both safety and security reasons, utilizing video analytics to detect humans on the lanes, large animals entering the tunnel, stopped vehicles, current traffic flow, etc. This particular install used video analytics from PureTech Systems which integrated into a Bosch Video Management System (BVMS).
The system has been installed for over a year and is working well.
2. Better to ceiling mount cameras 21' above road surface, or wall mount at a lower elevation?
There are several issues with tunnels, reflections and lighting require a video analytics product with good background modeling to avoid false alarms.
The other issue is overlapping targets. Advanced tracking does a good job separating targets based on various attributes, but mounting the cameras over the road gives the greatest separation of targets.
3. Is there practical benefit of using a thermal camera?
Thermal cameras help to eliminate some of the environmental effects of lighting and reflections. From a pure video analytics perspective, these would be preferred in tunnels. However, visible or low light cameras aid in gaining details or identification of targets. So, from a practical standpoint, we have found that a mix of sensor types is a good solution. For Tom Lantos we used MIC Series 612 Thermal and DINION cameras from Bosch.
4. Suggested distance between cameras along the roadway?
The answer to this question “depends” based on several factors:
Budget - Video analytics can detect out to great distances, but in a tunnel, the greater the distance to the target, the great the chance that the target is occluded by another car and cannot be discriminated. So, having more cameras closer together gives more target discrimination, but at a greater install cost.
Mounting Height – A greater mounting height allows for a greater detection distance with occlusion.
Camera Type – Use of thermal cameras or higher resolution cameras can increase this distance.
Lane congestion is also a consideration - If a single lane is not overly congested at any specific time, this allows for better target discrimination at greater distances.
As a guideline, we mounted thermal cameras (768x494) at 400m separation, using cameras viewing in both directions to accommodate for target overlap. You will likely need more cameras at each location to accommodate the width of the tunnel. That being said, you should really coordinate with the video analytics provider to determine a camera installation scenario that best balances budget and performance.
5. Possible surveillance interface with traffic management system?
Can you be more specific on this question? I’m not sure the type of interference you are referring to (Changing Lane Assignments based on time of day?)