Video stream and scene requirements for the Neural classifier

Hardware requirements for neural analytics operation

To configure the Object Presence Detection, do the following:

  1. Go to the Detection Tools tab.
  2. Below the required camera, click Create… → Category: Production Safety → Object Presence Detection.

By default, the detection tool is enabled and set to detect objects in the frame.

If necessary, you can change the detection tool parameters. The list of parameters is given in the table:

ParameterValueDescription
Object features
Record mask to archiveYesBy default, the sensitivity scale of the detection tool is recorded to the archive (see Displaying information from a detection tool (mask)). To disable the parameter, select the No value
No
Video streamMain streamIf the camera supports multistreaming, select the stream for which detection is needed. Selecting a low quality video stream reduces the load on the Server
Other
EnableYesThe detection tool is enabled by default. To disable the detection tool, select the No value
No
NameObject Presence DetectionEnter the detection tool name or leave the default name
Decoder modeAutoSelect a processing resource for decoding video streams. When you select a GPU, a stand-alone graphics card takes priority (when decoding with NVIDIA NVDEC chips). If there is no appropriate GPU, the decoding will use the Intel Quick Sync Video technology. Otherwise, CPU resources will be used for decoding
CPU
GPU
HuaweiNPU
Number of frames processed per second0.1Specify the number of frames that the detection tool will process per second. The value must be in the range [0.016; 100]
Selected object classes 

If necessary, specify the class of the detected object. If you want to display tracks of several classes, specify them separated by a comma with a space. For example, 1, 10.

    1. If you leave the field blank, the tracks of all available classes from the neural network will be displayed (Neural network file).
    2. If you specify a class/classes from the neural network, the tracks of the specified class/classes will be displayed (Neural network file).
    3. If you specify a class/classes from the neural network and a class/classes missing from the neural network, the tracks of a class/classes from the neural network will be displayed (Neural network file).

    4. If you specify a class/classes missing from the neural network, the tracks won’t be displayed (Neural network file)

TypeObject Presence DetectionName of the detection tool type (non-editable field)
Advanced settings
Neural network file

Specify the path to the neural network file

  • To train your neural network, contact AxxonSoft (see Data collection requirements for neural network training).
  • A trained neural network for a particular scene allows you to detect only objects of a certain type (for example, a person, a cyclist, a motorcyclist, and so on).
  • To ensure the correct operation of the neural network on Linux OS, the corresponding file must be located in the /opt/AxxonSoft/DetectorPack/NeuroSDK directory. 
Number of measurements in a row to trigger detection5

Specify the minimum number of frames on which the detection tool must detect an object to generate an event. The value must be in the range [5; 20]

Scanning modeYes

The parameter is disabled by default. To detect objects without changing the frame size, select the Yes value. To work in the scanning mode, the neural network must support the scanning mode

No
Basic settings
ModeCPU

Select a processor for the neural network operation (see Hardware requirements for neural analytics operation, Selecting Nvidia GPU when configuring detectors).

Nvidia GPU 0
Nvidia GPU 1
Nvidia GPU 2
Nvidia GPU 3
Intel GPU
Huawei NPU
Sensitivity 33

Specify the sensitivity of the detection tool empirically. The value must be in the range [1; 99]. The preview window displays the sensitivity scale of the detection tool that relates to the sensitivity parameter. If the scale is green, object isn't detected. If the scale is yellow, object is detected, but not enough to generate an event. If the scale is red, object is detected and the detection tool will generate an eevent, if the scale is red through the sampling period (50 seconds by default).


Example. The sensitivity parameter value of 40 means that the detection tool will generate an event when the scale has at least four divisions full over the entire detection period. An event will stop when the scale has less than two divisions full over the detection period. The detection tool will generate an event again if the scale has at least four divisions full over the entire detection period

By default, the entire frame is a detection area. In the preview window, you can specify the detection areas using the anchor points (see Configuring a detection area):

  1. Right-click in the preview window.
  2. If you want to specify the detection area by one or more rectangles, select Detection area (rectangle). If you specify a rectangular area, the detection tool will analyze only this area. The rest of the frame will be ignored.
  3. If you want to specify the detection area by one or more polygons, select Detection area (polygon). If you specify one or several polygonal areas, the detection tool will analyze the entire frame. The part of the frame not included in the specified polygons will be blacked out.

    You must select the detection area (polygon or rectangle) experimentally. For some neural networks the quality of detection will be better with rectangle, for others—with polygon.

  • For convenience of configuration, you can "freeze" the frame. Click the button. To cancel the action, click this button again.
  • To hide detection area, click the button. To cancel the action, click this button again.
  • To delete the selected area, click the button.

To save the parameters of the detection tool, click the Apply button. To cancel the changes, click the Cancel  button.