INTRADA HandHeld
|
Description | Features | Performance | Hardware |
HandHeld Recognition Performance
Obtaining good recognition performance is a real challenge when it comes to mobile license
plate reading. The quality of built-in cameras used by mobile devices is far inferior to
that of professional surveillance or traffic cameras. Moving vehicles or movements of the
operator during image capture blur the image. Direct sun or strong shadows result in
overexposed and underexposed images. Furthermore operators will point-and-click from different
points of view, resulting in a wide range of possible pan and tilt angles for the license plate
in the image.
Even under these extreme variations in image quality INTRADA HandHeld is capable of
delivering the requested performance. The three cases below illustrate the capabilities of
INTRADA HandHeld when dealing with different lighting conditions, variations in point of view,
and camera movement.
Lighting conditions
Lighting conditions

The cameras in handheld devices provide only limited compensation for changes in lighting
conditions. Direct sun or strong shadows very frequently result in overexposed and underexposed
images. INTRADA HandHeld software deals with these effects as the use of additional light sources (infrared,
flash) to compensate for this is not an option.
Point of view
Point of view

The efficient use on the road requires point-and-click operation without tedious aiming
procedures. This results in a wide variety in size, shape and location of the plate in
the acquired image. INTRADA Handheld is capable of handling all these variations and offers
the operator a flexible point of view.
Image quality
Image quality

Camera movements during image capture have a blurring effect. For single-handed
operated mobile devices movements are almost inevitable. Furthermore, many of the
handheld devices perform a built-in image compression. INTRADA Handheld processes
even these blurred and compressed images, making single shot operation possible.
INTRADA Handheld simply performs the task it is designed for: seamless
automatic license plate recognition.
|