View Source

To configure the Object tracker detector, do the following:

Go to the Detectors tab.
Below the required camera, click Create… → Category: Trackers → Object tracker.

By default, the detector is enabled and set to detect moving objects in the frame, on the basis of which their tracks are created.

Some parameters are set for all sub-detectors of the Object tracker simultaneously.

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

Parameter

Value

Description

Object features

Record objects tracking

Yes

By default, objects' tracks are recorded to the database. To disable the recording, select the No value

To obtain metadata, the video is decompressed and analyzed, which results in a heavy load on the server and limits the number of cameras that can be used on it.

No

Video stream

Main stream

For multi-stream cameras, select a stream for detection. Selecting a low-quality video stream reduces the server load

For the correct display of tracks in a multi-stream camera, all video streams must have the same aspect ratio.

Other

Enable

Yes

By default, the detector is enabled. To disable it, select the No value

No

Name

Object tracker

Enter the detector name or leave the default name

Decoder mode

Auto

Select a processor 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 are used for decoding

CPU

GPU

HuaweiNPU

Type

Object tracker

Name of the detector type (non-editable field)

Neural network filter

Enable filter

Yes

By default, the neural network filter is disabled. To filter out false tracks on a complex video image (foliage, glare, and so on), set the value to Yes (see Hardware requirements for neural analytics operation)

You can use a neural network filter either for the analysis of moving objects or for the analysis of abandoned objects. You cannot use two neural network filters simultaneously.

No

Moving object filter mode

CPU

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

If you specify another processor than the CPU, this device carries most of the computing load. However, the CPU is also used to run the detector.
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 Windows OS, Optimizing the operation of neural analytics on GPU in Linux OS).
Starting with Detector Pack 3.11, Intel HDDL and Intel NCS aren’t supported.
Starting with Detector Pack 3.14, Intel Multi-GPU and Intel GPU 0-3 are supported.

Nvidia GPU 0

Nvidia GPU 1

Nvidia GPU 2

Nvidia GPU 3

Intel NCS (not supported)

Intel HDDL (not supported)

Intel Multi-GPU

Intel GPU 0

Intel GPU 1

Intel GPU 2

Intel GPU 3

Huawei NPU

Abandoned object filter mode

CPU

Nvidia GPU 0

Nvidia GPU 1

Nvidia GPU 2

Nvidia GPU 3

Intel NCS (not supported)

Intel HDDL (not supported)

Intel Multi-GPU

Intel GPU 0

Intel GPU 1

Intel GPU 2

Intel GPU 3

Huawei NPU

Moving object filter file

Select the required neural network. To obtain a neural network, contact AxxonSoft technical support. If the neural network file isn't selected or selected incorrectly, the filter doesn't work

You cannot specify the network file in Windows OS. You must place the neural network file locally, that is, on the same server where you install Axxon One.
For correct neural network operation on Linux OS, place the corresponding file locally in the /opt/AxxonSoft/DetectorPack/NeuroSDK directory or in the network folder with the corresponding access rights.

Abandoned object filter file

Basic settings

Long-time abandoned object detection

Yes

By default, the parameter is disabled. To use the long-time abandoned object detection, select the Yes value. If you enable the Long-time abandoned object detection and the Enable filter parameters, it can reduce the number of false positives during detection

No

Abandoned object detection

Yes

By default, the parameter is disabled. To use the abandoned object detection, select the Yes value. Objects abandoned for 10 seconds or longer are detected

No

Max. object height

100

Enter the maximum height and width of the detected object as a percentage of the frame height/width. We recommend specifying the value so that it is slightly larger than the typical object in the image, taking into account its shadow. The detector doesn't generate an event if an object is larger or smaller. The value must be in the range [0.05, 100]

If you enable the Object calibration parameter in the tracker settings, then the maximum and minimum sizes of the objects are set in decimeters in the Leveling rod height parameter and not as a percentage of the frame size.

Max. object width

100

Alarm on object's max. idle time in area

60

Specify the time in seconds exceeding which the object is detected. This value must be in the range [15, 1800]

This parameter is used only for the Long-time abandoned object detection. We recommend selecting the parameter value starting from 15.

Min. object height

2

Enter the minimum height and width of the detected object as a percentage of the height/width of the frame. We recommend specifying the value so that it is slightly smaller than the typical object in the image. The detector doesn't generate an event if an object is larger or smaller. The value must be in the range [0.05, 100]

If you enable the Object calibration parameter in the tracker settings, then the maximum and minimum sizes of the objects are set in decimeters in the Leveling rod height parameter and not as a percentage of the frame size.

Min. object width

2

Motion detection sensitivity

25

Enter the sensitivity of the motion sub-detectors as a percentage. To detect objects with low contrast, the recommended sensitivity value is 35, for high-contrast objects, it is 15. The higher the sensitivity, the fewer changes in the frame are detected. The value must be in the range [0, 100]

Abandoned object detection sensitivity

9

Enter the sensitivity for abandoned object detection and long-time abandoned object detection as a percentage. The value must be in the range [0, 100]

The abandoned object detection sensitivity depends on the lighting conditions and is selected empirically. We recommend selecting the parameter value starting from 20. To detect objects with low contrast, the recommended sensitivity value is 35, for high-contrast objects, it is 15.

Advanced settings

Auto sensitivity

Yes

By default, the Auto sensitivity is enabled. We recommend enabling this parameter if the lighting changes significantly. To disable automatic adjustment of the sensitivity of the Object tracker sub-detectors, select the No value

We recommend enabling this parameter if the lighting changes significantly during the camera operation (for example, if the camera operates outdoors).

No

Track lifespan (starting with Detector Pack 3.14)

Yes

By default, the parameter is disabled. If you want to display the track lifespan for an object in seconds, select the Yes value

No

Leveling rod height

20

Enter the actual height of an object in decimeters that will be used as a reference (for example, the average height of a person). The value must be in the range [1, 100]

Frame size change

1280

By default, during the analysis, the frame is compressed to the specified size (by default, 1280 pixels on the larger side). The following algorithm is used:

If the original resolution on the larger side of the frame is greater than the resolution specified in the Frame size change parameter, then it is divided in half.
If the resulting resolution becomes less than the specified one, then the algorithm stops and the resulting resolution is used.
If the resulting resolution is still greater than the specified one, then it is divided in half again until it becomes smaller.

For example, the original video resolution is 2048x1536, the specified value is 1000. In this case, the original resolution is divided in half (512x384) twice because after the first division, the value on the larger side of the frame is greater than the specified value (1024 > 1000).
If detection errors occur when using a higher resolution stream, we recommend reducing compression.

Object calibration

Yes

By default, object calibration is disabled. To estimate the actual sizes of objects in a scene (for example, a height of a person) when analyzing a video image taking into account perspective distortions, select the Yes value

If you select the Yes value for the Object calibration parameter but don't set any leveling rod, the Object tracker is inoperable and doesn't generate tracks (track objects).
The correct Object calibration is impossible when you use cameras with optical distortions (for example, fisheye cameras).

No

Camera position

Wall

To filter out false events when using a fisheye camera, select the correct location of the device. The default value is Wall. This parameter isn't relevant for all other devices

Ceiling

Antishaker

Yes

By default, the parameter is disabled. To reduce camera shake, select the Yes value. We recommend using this parameter only when there is significant camera shake

No

By default, the entire frame is a detection area. If necessary, in the preview window, set:

one or more detection areas (see Configuring a detection area),
one or more ignore areas (see Configuring the ignore area).
visual size of objects (the No value is selected in the Object calibration parameter):
- Click the min button in the preview window. A rectangular area appears in the frame. Drag the anchor points to adjust the area to the minimum object size. The values update automatically in the Min. object height and Min. object width parameters.
- Click the max button in the preview window. A rectangular area appears in the frame. Drag the anchor points to adjust the area to the maximum object size. The values update automatically in the Max. object height and Max. object width parameters.

leveling rods (the Yes value is selected in the Object calibration parameter. The height of an object in decimeters is specified in the Leveling rod height parameter):

Click the button in the preview window.
In the window, click the mouse button.
From the list, select Leveling rod position.
Click the button in the preview window.

Specify visually the dimensions of the same object in different parts of the frame. To ensure accurate calibration, the following conditions must be met:

The leveling rod is set precisely according to a reference object (for example, the height of a person) standing in a given place in the scene.
You set at least three leveling rods in the frame.
All bottom points (bases) of the leveling rods are on the same physical plane (for example, the floor).
The bottom anchor points of any three leveling rods aren't positioned on the same line (horizontal, vertical, inclined, and so on).
Leveling rods are evenly distributed across the entire frame, covering the areas where objects are expected to appear. The larger the coverage area, the higher the accuracy.
Objects located at the same distance from the camera are marked in the frame with leveling rods of the same height.

For simple scenes, 3-5 leveling rods are set. For complex scenes with deep perspective, we recommend setting up to 10-15 leveling rods.

To create one leveling rod, add two anchor points by left-clicking in the frame:

The first point is the base of an object (for example, a person's feet).

The second point is the top of an object (for example, the top of a person's head).

To stop the video, click the button . Click this button again to resume playback.
To hide detection and/or ignore areas, click the button . Click this button again to display the areas.
To resize a leveling rod, click and hold down the anchor point and drag it in the required direction.
You can move the leveling rod by drag-and-dropping it.
To delete a created leveling rod or area, click the button.

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

Configuration of the Object tracker detector is complete. If necessary, you can create and configure the necessary sub-detectors on the basis of the Object tracker (see Abandoned object, Standard sub-detectors). General parameters are set for all its sub-detectors.