Lum wrote:
ed_m wrote:
ah here's the paragraph of interest in that respect:
Quote:
The time that the vehicle has taken to drive through the monitored zone is compared against a baseline measurement. UK Home Office guidleines specify that the baseline measurement should be the shortest route that a car should take (duh

); this measurement is carried out using a calibrated (against what?) satellite positioning system approved by the home office
It makes sense in a way. Believe it or not GPS is accurate for this purpose. Sure GPS can drift by up to 40 feet, but the drift is constant across a large area, so unless you're running SPECS down a city centre street with tall buildings then you can accurately get the distance between cameras by this method.
Yup i'm happy GPS offsets drift relatively slowly, say over several minutes, enough to grab a log of one route option. For a more complex network of roads i expect they'd be logging over several hours, so stitching the data together may not be totally trivial. Unless they're going to run one calibration run for every entry/exit combination for a more complex system would soon mount up.
Not sure why they need to spec a GPS, I'd have thought a calibrated survey vehicle was just as easy and would provide a better paper trail.
I'd be even happier if i knew the approved GPS system was DGPS based, even better with a fixed base station at one of the entry/exit points.
Lum wrote:
Does it say anything about how each individual camera gets it's time (again GPS would be valid for this) or are they just dumb cameras all linked to a central controller somewhere, thus only one time source to worry about.
The articles are only 2 x a few pages so not in great depth.
The redspeed system refers to the ANPR data being transferred with "two seperate GPS time stamps for added accuracy".
I'm not clear where the processing takes place in specs, the impression i get is at the camera end or at least roadside where "each camera unit is a remote node, communicating back to a central server using wired or wireless telecommunications".
The redspeed article gives a clear structure with a laser triggered image capture, to a "remote image detector" to do the ANPR, sending the data & timestamps to a "front office decision making unit where the information is held pending the vehicle leaving the monitored zone".
(and to counter the next question, if no data or match is found the information is deleted)