Vehicle Motion Detection

The motion detection module lets you aquire vehicles which may have no plate or a badly damaged plate.

The configuration of the motion detection engine needs some measure of adjustment to get it running at its optimum for the particular highway being monitored, but once setup accurately it should identity all passing vehicles in the field of view. Due to the type of motion detection engaged, it is extremely resiliant to false triggering from, for example, moving shadows and multiple triggers from long vehicles.

To use this module, you must facilitate 'Transmit Blank Plates' from the current lane’s settings and the triggered ALPR must be deactivated.

The motion detection works autonomously for each ALPR camera (it cannot be configured on surveillance cameras). Every camera can be configured with individual settings, so each camera's exact sight of the road is being observed.

Traditional motion detection operates by comparing sections of a video, searching for variations between consecutive frames. The motion detection process works by seeking horizontal and vertical lines, which are extremely specific for every vehicle and see how these alter from frame to frame.

The "motion detection area" is defined by the green rectangle displayed in the video window. The four corners of the rectangle can be pulled about to include the necessary area to use for motion detection.

The white gridlines split the motion detection area into smaller boxes. The number of boxes can be arranged by modifying the Number of rows and Number of columns values. For each frame, the motion detection algorithm determines the number of horizontal and vertical lines there are within each box, which are displayed as red and green lines within the video window.

Once the configured criteria have been reached, the current frame is shown in the smaller "motion detection result" window. If the ALPR engine is unable to identify a plate, for whatever reason, during a motion-detected trigger phase, this is the image that will be transmitted as a "blank plate".

Configuration Criteria Settings:

No. Boxes Hit

Stipulates the number of cells within the main box that must either include lines or display variation between frames prior to the motion event is triggered. The greatest value authorized is defined as the total number of white boxes inside the main green rectangle.

No. Lines Per Box

How many lines each box must hold before being deemed to contain lines. This helps to ignore any static lines in the image from, for example, road markings.

Hold Trigger for n Frames

Once a motion event is triggered, this rule stipuates for how long, in frames, the trigger lasts. This module will stop multiple triggers happening from the front/rear of the vehicle (depending on traffic direction).

Number of Rows

Number of rows in the main box.

Number of Columns

Number of columns in the main box.

Priorities

This control sets whether the horizontal or vertical lines in each cell are used to work out the "No. Boxes Hit". This helps to distinguish between the front and back of a vehicle. For example, consider vehicles moving from the top to the bottom of the field of view: The front of the vehicle will have many more vertical lines (from radiator grill, etc) than the back edge. If a long vehicle goes by, the front will trigger the motion event. As the vehicle passes, the rear will not set off an event because it has a smaller number of vertical lines.

Look For

Tells the motion detection algorithm whether to count the number of cells that currently contain lines or whether to count the number of cells that changed between consecutive frames.