LPR applications can be split into three categories: Access Control (AC), Traffic Law Enforcement (LE), and Road Patrol (RP).

Access Control Traffic Law Enforcement Road Patrol Application Parameters

    Access control refers to the cases that a vehicle passes a fixed passage at a reduced speed or with a full stop, such as at a toll station or the entrance/exit of a region. In access control scenarios, the camera is often placed less than 5 meters from the plate, within -30 to +30 degrees in pan and 0 to 60 degrees in tilt (as 0 degree tilt is parallel to the ground). In the image, the width of a plate is between 0.2 to 0.25 the width of the image (shown as width ratio in Application Parameters), and its orientation is less than 10 degrees (Note that both are measured in the image by the plate projected onto the image plane). The parameters in Application Parameters are generalized from the 681 images collected at various access control scenes. The illumination covers indoor, outdoor, daytime, night time, and different weather conditions. If measured by the average intensity over a plate, it varies from 60 to 130 degrees in an 8-bit gray scale. The above parameters of access control are summarized in Application Parameters, along with those for the other two applications.

    Traffic law enforcement refers to the cases that a vehicle travels at a regular or higher speed but violates traffic laws, such as a traffic signal or speed limit, and is captured by a roadside camera. 757 images were collected for this application category.

    Road patrol refers to the cases that the camera is installed or handheld on a patrolling vehicle which takes images of the vehicles with arbitrary viewpoints and distances. The purposes of road patrol include search for lost vehicles, scan for parking violation, security check in a restricted area, or others. 611 images were collected for this application scenario.

    Application Parameters. Variable and variation scopes for three application, generalized from the AOLP database. The bottom row give the number of sample in each subset.

    It can be seen from Application Parameters that AC is with variables of narrower ranges of variation, while RP is with variables of wider ranges of variation. The latter is expected to take a longer time for plate detection as it comes with a larger search space to go through in the plate detection phase. Although LE and RP have a few variables with the same or similar ranges of variation, the most significant difference between the two is on the size, pan and orientation. Since the LE samples were mostly collected from cameras installed on road sides with a constant distance and viewing angle, the variation on size, pan and orientation is smaller than those collected by mobile cameras in the RP scenarios.

    This AOLP database was created by Gee-Sern Hsu, Addis Chen, Ming-Hong Chen and Pendry Alexandra, members of the Artificial Vision Laboratory,Department of Mechanical Engineering,National Taiwan University of Science and Technology. See this paper for details.