Documentation for Auto Intellect 5.1. Documentation for other versions of Intellect is [available too].

Previous page Next page

The following video camera requirements should be met in order to ensure the recognition of license plate numbers using the Auto-Intellect system.

  1. Size of each image in sequence is not less than 320 dpi in width and not less than 240 dpi in height.
  2. Height of LP symbols on the image is not less than 16px (noninterlaced scan) or 32px (interlaced scan).
  3. License plate consists of latin symbols and digits.
  4. Brightness of field of LP symbols is changed not more than on 15 units (from 255 gradations);
  5. Brightless of field of license plate is chsnged not more than on 30 units (from 255 gradations);
  6. Only single-line plates can be recognized;
  7. Difference of symbol sizes on the vehicle direction in the frame is not more than 3 units.
  8. Each license plate is to be presented in the input data not less than in 3 frames.

Basic characteristics of used video cameras are presented in the General requirements for mounting and configuring of cameras section.

Note.

Recommended video stream speed for the AxxonSoft software module is 3 fps, i.e. the license plate should be entirely displayed and read at least in three frames for the correct LP recognition and determining the directions of vehicle moving.

Main parameters required for mounting video camera for working with the Axxonsoft software module are presented in the following table.

Parameter

Maximally allowed value

Vertical angle of video camera excursion to plate of road

30°

Angle of horizontal video camera excursion from direction of vehicle moving

20°

Angle of license plate heel to the bottom border of frame

5°

The closer specified parameters to null, the higher probability of recognition.

Maximal values of excursion are presented in the following figure.