To configure smoke (fire) detection tool:

1. If a camera supports multistreaming, select the stream for which detection is needed (1). Selecting a low-quality video stream allows reducing the load on the Server.
   
2. Select a processing resource for decoding video streams (2). When you select a GPU, a stand-alone graphics card takes priority (when decoding with NVidia NVDEC chips). If there's no appropriate GPU, the decoding will use the Intel Quick Sync Video technology. Otherwise, CPU resources will be used for decoding.

3. Set the interval between the processes frames in seconds (3). 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.

   
   

4. Select the processor for the neural network — CPU, one of GPUs or a IntelNCS (4). 

   It may take several minutes to launch the algorithm on an NVIDIA GPU after you apply the settings.

   
   

   If you specify other processing resource than the CPU, this device will carry the most of computing load. However, the detection tool will consume CPU as well.

   
   

5. Select a neural network file (5).  The following standard neural networks for different processor types are located in C:\Program Files\Common Files\AxxonSoft\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

   Enter full path to a custom neural network file into this field. This is not required if you use standard neural networks which are selected automatically.

   For correct neural network operation under Linux, place the corresponding file in the /opt/AxxonSoft/AxxonNext/ directory.

   
   

6. Set the minimum number of frames with smoke (fire) for triggering the tool (6). The value should be in the [5;20] range.
7. You can experiment with the sensitivity of the tool (7).  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).
      

8. By default, the detection is performed over full image area. In the preview window, you can set several detection zones by their anchor points as follows:

   1. Right-click anywhere in the Preview window.
      

   2. Select Detection area (rectangle) for a rectangular zone. If you specify a rectangular area, the detection tool will work only within its limits; the rest of the FOV will be ignored.

   3. Select Analytics Area (polygon) to set one or several polygonal zones. If you specify one or several polygonal areas, the detection tool will process the entire FOV while the remaining part of the FOV will be blacked out.

      You can configure detection zones similarly to privacy masks in scene analytics (see Setting General Zones for Scene Analytics).