Iteris VersiCam, Руководство пользователя

Страница: 31 / 84

User Guide
31
4. SYSTEM CONFIGURATION
Vantage Video Detection Zone Theory 101
You may have heard the term "complex edges" associated with the Vantage video detection systems.
Vantage systems utilize complex edge detection as one of the prominent techniques to effectively detect
vehicles. Vehicles tend to have more complex edge content than other undesirable artifacts in the field of
view. This edge based algorithm gives Vantage video detection products their exceptional detection stability,
even under severe environmental conditions.
For increased opportunity to detect, zones can also be overlapped from front to back or from side to side.
This will compensate for vehicles that may not stop exactly at the stop line or a vehicle traversing lanes in the
detection area.
Appropriate zone sizing rules should still be carefully observed. The best policy is to use the least number of
zones possible to obtain the desired operation and performance. Some people think if one overlapped zone
is good, then ten overlapped zones must be better. This is not necessarily the case, because increased
opportunity to detect is non-discriminatory. This means that an increased opportunity to detect vehicles may
also increase the opportunity to detect undesirable artifacts. It is therefore a wise policy to avoid excessive
overlapping, especially over a troublesome area since increased opportunity to detect may lead to false calls
under some circumstances.
Lane Structure
Background: Starting with firmware revision 08.01.15SP1 VersiCam processors take advantage of a lane zone
structure. The lane zone structure tracks vehicles as they travel through the zones. The lane structure helps
to detect vehicles better and reject false calls due to shadows. The lane structure extends in front and behind
of the existing zones for greater tracking coverage.
Guidelines for setting up effective lane zones: Two or more zones in the same lane are needed to make a lane
structure. A single zone will not be included in a lane structure. The zones need to be assigned an output
channel. If a zone’s output channel is assigned to none then it will not be included in a lane structure.

Important:
If a second detection zone is not required for intersection control it must still be drawn
to create the lane structure. The second zone should have a valid detector output, for example Proc:1
and the zone type should be set to none. This will ensure system performance will be maximized.