Page History

Configuration of the Neurocounter module includes: configuring the detection tool, detector and selecting the area of interest. You can configure the Neurocounter module on the settings panel of the Neurocounter object created on the basis of the Camera object on the Hardware tab of the System settings dialog window.

Image Modified

Configuring the

...

detector

...

1. Go to the settings panel of the Neurocounter object.
   Image Removed
2. Set the Show objects on image
 checkbox to
3. checkbox to frame the detected
objects
4. object on the video image in the debug window (see Start the debug window).
   Image Added

The Main settings tab

You can configure the neurocounter main settings on the tab of the same name.

1. From the Camera position drop-down list, select:
   1. Wall—objects are detected only if their lower part gets into the area of interest specified in the
detection tool
2. 1. detector settings.
   2. Ceiling—objects are detected even if their lower part doesn't get into the area of interest specified in the
detection tool
3. 1. detector settings.
4. In the Number of frames for analysis and output field, specify the number of frames to be processed to determine the number of objects on them.
5. In the Frames processed per second [0.016, 100] field, specify the number of frames processed per second by the neural network in the range from 0.016 to 100. For all other frames interpolation will be performed—finding intermediate values by the available discrete set of its known values. The greater the value of the parameter, the more accurate the detection tool operation, but the higher the load on the processor.
6. From the Send event drop-down list, select the condition by which an event with the number of detected objects will be generated:
   • If threshold exceeded is triggered if the number of detected objects in the image is greater than the value specified in the Alarm objects count field.
   • If threshold not reached is triggered if the number of detected objects in the image is less than the value specified in the Alarm objects count field.
   • On count change is triggered every time the number of detected objects changes.
   • By period is triggered by a time period:
     1. In the Event periodicity field, specify the time after which the event with the number of detected objects will be generated.
     2. From the Time interval drop-down list, select the time unit of the counter period: seconds, minutes, hours, days.
7. In the Alarm objects count field, specify the threshold number of detected objects in the area of interest. It is used in the If threshold exceeded and If threshold not reached conditions. The default value is 5.

8. In the Recognition threshold [0, 100] field, enter the neurocounter sensitivity—integer value from 0 to 100. The default value is 30.

   Info
   title Note
   The neurocounter sensitivity is determined experimentally. The lower the sensitivity, the higher the probability of false alarms. The higher the sensitivity, the lower the probability of false alarms, however, some useful tracks can be skipped (see Example of configuring Neurocounter for solving typical task).

9. From the Device drop-down list, select the device on which the neural network operates: CPU, one of NVIDIA GPUs, or one of Intel GPUs. Auto (default value)—the device is selected automatically: NVIDIA GPU gets the highest priority, followed by Intel GPU, then CPU.
   

   Note
   title Attention!
   We recommend using the GPU. It can take several minutes to launch the algorithm on the NVIDIA GPU after you apply the settings. You can use caching to speed up future launches (see Optimizing the operation of neural analytics on GPU). In the Detector Pack subsystem of version 2.0, the support of Intel HDDL is removed, therefore, when you update from version 1.0, the Not supported option is selected automatically instead of this device variant, and detectors don't operate. To resume the detector operation, select the required device from the list.

10. From the Object type drop-down list, select the object type:
    • Human—the camera is directed at a person at an angle of 100-160°;
    • Human (top-down view)—the camera is directed at a person from above at a slight angle;
    • Vehicle—the camera is directed at a vehicle at an angle of 100-160°;
    • Person and vehicle (Nano)—person and vehicle recognition, small neural network size;
    • Person and vehicle (Medium)—person and vehicle recognition, medium neural network size;
    • Person and vehicle (Large)—person and vehicle recognition, large neural network size;
    • Human top-down view (Nano)—the camera is directed at a person from above at a slight angle, small neural network size;
    • Human top-down view (Medium)—the camera is directed at a person from above at a slight angle, medium neural network size;
    • Human top-down view (Large)—the camera is directed at a person from above at a slight angle, large neural network size;
      

      Info
      title Note
      Neural networks are named taking into account the objects they detect. The names can include the size of the neural network (Nano, Medium, Large), which indicates the amount of consumed resources. The larger the neural network, the higher the accuracy of object recognition.

Set the Scanning mode checkbox to detect small objects. If you enable this mode, the load on the system increases. So we recommend specifying a small number of frames processed per second in the Frames processed per second [0.016, 100] field. By default, the checkbox is clear. For more information on the scanning mode, see Configuring the Scanning mode.
12. By default, the standard (default) neural network is initialized according to the object type selected in
the Object type drop-down list
13. step 14 and the device
selected in the Device drop-down list
14. type in step 13. The standard neural networks for different processor
types 
15. types are selected automatically; you must not do it manually. If you use a custom neural network, then click the
Image Removed
16. Image Added button to the right of the Tracking model field, and in the standard Windows Explorer window that opens, specify
the path to the
17. its file
.
18. 
    

    Note
    title Attention!
    To train a neural network, contact

the

19. AxxonSoft technical support (

see

20. see Data collection requirements for neural network training). A neural network trained for a specific scene allows

you to detect

21. detecting objects of a certain type only (for example, a person, cyclist, motorcyclist, and so on).

22. Set the Scanning mode checkbox to detect small objects. If you enable this mode, the load on the system increases. So in step 10 we recommend specifying a small number of frames processed per second. By default, the checkbox is cleared. For more information on the scanning mode, see Configuring the Scanning mode.
23. Set the Model quantization checkbox to enable themodel quantization. By default, the checkbox is
clear
24. cleared. This parameter allows
you to reduce
25. reducing the consumption of the GPU
processing
26. 's computational power.
    

    Info
    title Note
    AxxonSoft conducted a study in which a neural network model was trained to identify the characteristics of the detected object. The following results of the study were obtained: model quantization can lead to both an increase in the percentage of recognition and a decrease. This is due to the generalization of the mathematical model. The difference in detection percentage ranges within ±1.5%, and the difference in object identification ranges within ±2%. Model quantization is only applicable for NVIDIA GPUs. The first launch of

a detection tool with quantization enabled may is used

27. the detector with the activated quantization feature can take longer than a standard launch. If you use the GPU caching

 

28. , next time

a detection tool

29. the detector with quantization

will run

30. runs without

delay

31. delays.

32. If necessary, specify the class of the detected object in the Target classes field. If you want to count and display tracks of several classes, specify them separated by a comma with a space. For example,
 
33. 1,
 
34. 10.
    The numerical values of classes for the embedded neural networks: 1—Human/Human (top view), 10—Vehicle.
    

    Info
    title Note
    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

35. are counted and displayed (Object type, Neural network file). If you specify a class/classes missing from the neural network, tracks

won

36. aren't

be

37. counted and displayed.

38. From the Device drop-down list, select the device on which the neural network will operate: CPU, one of NVIDIA GPUs, or one of Intel GPUs. Auto (default value)—the device is selected automatically: NVIDIA GPU gets the highest priority, followed by Intel GPU, then CPU.
    

    Note
    title Attention!
    We recommend using the GPU. It may take several minutes to launch the algorithm on NVIDIA GPU after you apply the settings. You can use caching to speed up future launches (see Optimizing the operation of neural analytics on GPU).

From the Object type drop-down list, select the object type:
39. Human—the camera is directed at a person at the angle of 100-160°.
40. Human (top-down view)—the camera is directed at a person from above at a slight angle.
41. Vehicle—the camera is directed at a vehicle at the angle of 100-160°;
42. Person and vehicle (Nano)—person and vehicle recognition, small neural network size;
43. Person and vehicle (Medium)—person and vehicle recognition, medium neural network size;
Person and vehicle (Large)—person and vehicle recognition, large neural network size
44. In the Recognition threshold [0, 100] field, enter the neurocounter sensitivity—an integer value in the range from 0 to 100. The default value is 30.
    

    Info
    title Note
    The neurocounter sensitivity is determined experimentally. The lower the sensitivity, the higher the probability of false alarms. The higher the sensitivity, the lower the probability of false alarms, however, some useful tracks can be skipped. See Example of configuring neurocounter for solving typical tasks.

45. In the Number of frames for analysis and output field, specify the number of frames that must be processed to determine the number of objects in them.
46. In the Frames processed per second [0.016, 100] field, specify the number of frames processed per second by the neural network in the range from 0.016 to 100. For all other frames, interpolation is performed—finding intermediate values by the available discrete set of its known values. The greater the value of the parameter, the more accurate the detector operation, but the higher the load on the processor.

The Additional settings tab

You can configure the nerocounter additional settings on the tab of the same name.
Image Added

1. From the Send event drop-down list, select the condition by which an event with the number of detected objects is generated:
   • If threshold exceeded is generated if the number of detected objects in the image is greater than the value specified in the Alarm objects count field;
   • If threshold not reached is generated if the number of detected objects in the image is less than the value specified in the Alarm objects count field;
   • On count change is generated every time the number of detected objects changes;
   • By period is generated by the time period:
     1. From the Time interval drop-down list, select the time unit of the counter period: seconds, minutes, hours, days
     2. In the Event periodicity field, specify the time after which the event with the number of detected objects is generated. The range of values: from 1 to 100—for seconds, minutes, hours; from 1 to 20—for days.
        

        Info
        title Note

Neural networks are named taking into account the objects they detect. The names can include the size of the neural network (Nano, Medium, Large), which indicates the amount of consumed resources. The larger the neural network, the higher the accuracy of object recognition

2. • 1. If the entered value exceeds the allowable range, then after you click the Apply button, the maximum value is set automatically.

3. In the Alarm objects count field, specify the threshold number of detected objects in the area of interest. It is used in the If threshold exceeded and If threshold not reached conditions. The default value is 5.

Selecting the area of interest

1. Click the Settings button.
The 
2. As a result, the Detection settings window opens.
   
Image Removed
3. Image Added
   
4. Click the Stop video button (1)
to
5.  to pause the playback and capture the frame of the video image.
6. Click the Area of interest button (2) to specify the area of interest. The button
will be
7. is highlighted in blue.
   
Image Removed
8. Image Added
   
9. On the captured frame, sequentially set the anchor points of the area in which the objects
will be detected
10. are detected by using the mouse (3). The rest of the frame
will be
11. is faded. If you don't specify the area of interest, the entire frame is analyzed.

    Info
    title Note
    You can add only one area of interest. If you try to add a second area, the first one

will be

12. is deleted. To delete an area, click the

Image Removed

13. Image Added button to the right of the Area of interest button.

14. Click the OK
 button to close the Detection settings window
15. button (4) to save the detector settings and return to the settings panel of the Neurocounter object.
16. Click the Apply button
to save the changes
17. .

Configuring the Neurocounter module is complete.