Page 41/89
VIDAR
User Manual
Adaptive Recognition
America
Adaptive Recognition
Hungary
Adaptive Recognition
Nordic
Adaptive Recognition
Singapore
WWW.ADAPTIVERECOGNITION.COM
7.
EVENTS AND ANPR
This chapter explains the mechanism of analysing the video stream, segmenting it to image
sequences corresponding to one passing vehicle, then finding, validating, storing and publishing the
license plate of said vehicle.
The basic unit of the process is an event, typically consisting of one passing vehicle. In order to
segment the video stream into events, an external or internal trigger is needed. A trigger is a pair of
time coordinates: a start time stamp and an end time stamp, representing the duration of the event.
An internal trigger source may be:
•
Motion Detection: Image processing algorithm designed to detect motion in a sequence of
images
•
Scheduler: Regular series of triggers, for example every 5 seconds
•
Plate Finder Trigger: Image processing algorithm designed to detect a license plate inside of
a predefined ROI.
•
Laser Trigger (optional): uses distance measurements taken by the built-in laser device
An external trigger source may be:
•
GPI: A TTL level input to the camera (not all models)
•
UART: A UART (RS232 or RS485 level) input to the camera (not all models)
•
SW: Direct software-based trigger
The above trigger options cover a wide range of applications, for example:
•
Induction loop linked to the GP input
•
Radar communicating through the serial port of the camera
•
A laser software trigger via Ethernet.
It is possible to combine multiple triggers. An example: license plate recognition is required only
during a certain hour of the day: combine Vehicle Detection with an instance of Scheduler set to be
asserted during the required hour. This feature is configured in the Event Manager page.
Summary of Contents for Vidar
Page 1: ......