Documentation for Axxon One 2.0. Documentation for other versions of Axxon One is available too.

Previous page Camera requirements for the License plate recognition RR, Vehicle recognition RR and their sub-detectors  Optimizing the operation of License plate recognition RR and Vehicle recognition RR on GPU in Windows OS Next page

Recommendations

  1. You must install the camera according to the requirements (see Camera requirements for the License plate recognition RR, Vehicle recognition RR and their sub-detectors).

  2. You must select an area in the frame that meets the following conditions:

    1. The sides of the selected area (Speed detection area height and Speed detection area width) must be equal in pairs.

    2. The angles of the selected area must be 90° in the real world.

  3. When you specify the area in the Surveillance window, you must focus on:
    1. Traffic lanes (solid and broken lines on the roadway).
    2. Road curbs.
    3. Vehicle bumpers and wheel contact points with the roadway.
    4. Crosswalks.
  4. The quality of vehicle recognition will be higher if the selected area occupies a significant part of the frame and is as close to the camera as possible.

  5. You can use the dimensions of a particular vehicle as a reference point. For example, Hyundai Solaris: length (Speed detection area heigh)—2.6 meters, width (Speed detection area width)—1.73 meters

Attention!

Margin of error in determining the speed of vehicles is about 5% with a good calibration, as this function gives only a general idea of the situation. The consumption of resources doesn't change when the Speed detection parameter is enabled.

Examples of the corresponding selected area

 Lines are parallel to road marking and vehicle bumpers:


  • No labels