I'll give it your question a try...but Craig is pretty knowledgeable of such things.
I believe today's red light camera systems a bit more sophisticated (in light of previous court challenges, I suspect) and I believe they provide still and or video documentation subsequent to (and preceding) the alleged violation.
With speed-on-green technology (which is already built-in to these systems but may not be turned-on) induction strips and/or infrared laser (lidar) can be used to detect a (rapidly) approaching vehicle and their algos can "predict" a likely violation based on time metrics and distance/speed approaches.
When Roy and I were involved in the intense FireBall Run, last year, I actually made an aggressive approach to an intersection followed by very hard braking which resulted in a false-trigger of a red light camera monitored intersection in Houston.
As it turned out, we didn't actually "run the red light," but I was fully prepared to fight a ticket, if it has appeared in my mailbox.
Fortunately, it did not.
I believe that's what the pre- and post- analysis is used for--to add a more accurate context for the photographed event.
I hope this helps you, but Craig may be better suited to give you a more thorough answer than I.
With respect to the banner, thanks, I believe we have something that I can make available for you...thanks for thinking of us.