Versions Compared

Key

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

...

  1. Use color cameras. With black and white cameras, the detection quality can be noticeably worse.
  2. The video resolution must be at least 640x360.
  3. FPS directly affects detection efficiency; the response rate of the detection module depends on the number of frames per second, by default, it is 1 per every 10 seconds (in most cases, the default value of 0.1 is sufficient it is enough to detect smoke or fire which develops in a fire/smoke that lasts more than one 1 minute).

  4. Smoke or fire must be visually separated from the background.

  5. The area of smoke / fire must

    The minimum area of fire/smoke in the frame depends on the neural network used. For a standard neural network (see Configuring Smoke and Fire Detection Tools), the area of fire/smoke should be at least 10% of the

    area of the frame

    frame. 

    Info
    titleNote

    In some cases, when the fire is well contained, it may be sufficient for a detection tool that the width/height of the fire zone is 1-3% of the width/height of the frame.


  6. Fire/smoke should be visually separated from the background.
Note
titleAttention!

If a rectangular detection area is specified on the video image, then the above requirements are relevant for it, and not for the entire frame (see Configuring Smoke and Fire Detection Tools).


Tip

Hardware requirements for neural analytics operation

...