Below we have provided a basic instruction on how to configure license plate recognition functionality of AutoCode LPR module. You can find detailed description of recognition settings in AutoCode administrator's guide.
- Create a new AutoCode recognition module object on the Hardware tab.
Select a Camera object (which will provide a video for LPR module to analyse) on the settings pane of the newly created object.
Info When testing, you might probably select a Camera object which receives video frames from the local file system.
Click on the Country list button to select appropriate license plate templates in the window that appears.
- Click on Recognition button to open Recognition settings window.
Specify a so-called "recognition zone", a zone in the image where the license plate detection will be done.
Info The rest of the image (which is not included into the recognition zone) is ignored by the recognition engine. By default, if recognition zone is not specified, license plates are searched on the whole frame.
Specify maximum and minimum sizes of license plate images in the frame. You can do it either graphically or by inputting the exact values (in pixels).
Info Maximum size of license plate's image. If the license plate image's size exceeds this value, the recognition engine will ignore this license plate.
Minimum size of license plate's image. If the license plate image's size is less than this value, the recognition engine will ignore this license plate.
Select either Parking or Road license plate recognition mode.
Info Depending on selected mode, the recognition engine's advanced settings receive values which are optimized for specific traffic surveillance conditions (i.e. parking or road).
Set the Thread count parameter.
Info Number of system threads created/used for analysis of this video stream is an advanced setting of the recognition engine. You will find its detailed description in AutoCode administrator's guide. As a rule of thumb, you can assign it a value which equals a number of traffic lanes in the video.
