Requirements for video parameters when working with fire and smoke detection


Unlike standard smoke / fire detection systems, smoke and fire software detection tools face many issues with the scene and the background mage. We cannot  warrant 100% smoke / fire detection. The smoke and fire detection tools are meant to increase the likelihood of detecting smoke / fire. However, there may be both false alarms and failures to detect actual fire / smoke events in the camera's FoV.

We keep improving  smoke and fire detection and use machine learning based on a Neural network.

If the fire / smoke detection tools does not respond to actual fire / smoke events, please record a video clip and send it to Grundig. We will update C-Werk to refine detection. Help us train the neural network with video feeds from you scene to deliver best results for your fire security.

To configure smoke (fire) detection tool:

  1. Set the interval between the processes frames in seconds (1). The value should be in the [1;30] range.

    The default values (5 frames and 10 seconds) indicate that the tool will analyze one frame every 10 seconds.  When smoke (fire) is detected in 5 frames, the tool will trigger.


  2. Select the processor for the neural network — CPU, one of GPUs or a IntelNCS (2). 

  3. Select a neural network file (3). The following standard neural networks for different processor types are located in C:\Program Files\Common Files\Video Management System\DetectorPack\NeuroSDK:

    smoke_movidius.ann

    Smoke detector / IntelNCS

    smoke_openvino.ann

    Smoke detector / CPU

    smoke_original.ann

    Smoke detector / GPU

    fire_movidius.ann

    Fire detector / IntelNCS

    fire_openvino.ann

    Fire detector / CPU

    fire_original.ann

    Fire detector / GPU


  4. Set the minimum number of frames with smoke (fire) for triggering the tool (4). The value should be in the [5;20] range.
  5. You can experiment with the sensitivity of the tool (5). The value must be in the range [1; 99]. The preview window displays the sensitivity scale of the detection tool. It is color-coded as follows:

    1. Green - smoke (fire) not detected.  

    2. Yellow - smoke (fire) detected, but not enough to trigger the tool.

    3. Red - smoke (fire) detected.

      If you increase the sensitivity value, you have more alerts (red scale).

  6. In the preview window, you can set the detection zones with the help of anchor points much like privacy masks in Scene Analytics (see Setting general zones for scene analytics). By default, the entire FoV is a detection zone.

In some cases, setting the detection zone can improve the quality of smoke / fire recognition.