Versions Compared

Old Version 1

changes.mady.by.user Alena Kniazeva

Saved on

compared with

New Version Current

changes.mady.by.user Bahdan Tsiulkou

Saved on

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

Camera The requirements for the cameras when you work with the Crowd detection (TVNVA) module  module are given in the following table.

:: detection tool performance may may not correctly is recommended, : may not correctly e.g. may not correctly recording : may not correctly
Camera
  • Resolution: from 265x256 pixels and higher

Lighting

  • The best
  • detector operation is achieved under good lighting: no flash, with even lighting, in normal weather conditions (no rain). In conditions of insufficient lighting (for example, at night), the quality of algorithm performance
  • can decrease.

  • Analytics

  • can work

  • improperly with light and shade variations on objects: we recommend the homogeneous lighting

  • without complex light-refracting forms

Scene and camera angle
  • Analytics
  • can work
  • improperly if the scene is a repeating pattern (
  • for example, a crosswalk and a cart on it).
  • Analytics
  • can work
  • improperly when
  • you record a close-up view with only one face in it
Object images
  • Analytics
  • can work
  • improperly if there are objects that look like people (monuments, animals) in the camera FOV
Note
titleAttention!

We don't guarantee the correct detector operation if you use the video stream or frames that were masked before they got to Axxon PSIM by the IP camera means or similar methods.