Red-Light Camera Ticket Revenue and Short Yellows

"A Fort Meyers news station reports a nerdy husband getting his wife out of a red-light camera ticket by proving the light was set with too short of a yellow. Then he goes out and proves that nearly 90% of the lights are set an average of about 20% too short. Is this a local incident, or have local governments nationwide found a new revenue source? What puzzles me is how a single picture can tell if you ran a light. If you are in the intersection before the light turns red, you have not run it, even if it takes a little while to clear it (say to yield to an unexpected obstacle). Wouldn't you need two pictures — one just before the light went red showing you are not in the intersection, and another after the light went red showing you in the intersection?"

http://yro.slashdot.org/story/10/04/12/225248/Red-Light-Camera-Ticket-Revenue-and-Short-Yellows