Loading...
The Face Tracker Tool tracks a detected face/ body or eyes and outputs its position data via the Facetracking Input Node for further processing.
Open the Face Tracker Tool from the Devices menu. In the Face Tracker Tool you set up and connect to the unit/stream. Once connected you see the video input displayed in the preview and you can start setting up your tracking.
Choose your DirectShow Camera Device, NDI- or RTSP stream from the drop-down list.
Select the format (resolution@frame rate, aspect ratio) in the drop-down list. Please note that the format is automatically set when using NDI- or RTSP streams.
Enter an URL in case you chose "RTSP" from the Device list. This can be an address from a web cam within your local network for example or a link to a public stream.
Choose the Tracking Method from the drop-down list. The following modes are available:
Eye
Eye Tree Eyeglasses
Frontal Face Alt
Frontal Face Alt2
Frontal Face Alt_Tree
Full Body
Lower Body
Profile Face
Upper Body
This parameter specifies how much the image size is reduced at each image scale.
This parameter specifies how many neighbors each candidate circle should have to retain it. This parameter will affect the quality of the detected parts. Higher value results in less detections but with higher quality.
Enter the minimum possible object size in percent. Objects smaller than are ignored.
Enter the maximum possible object size in percent. Objects larger than are ignored.
Activates/ Deactivates the preview for displaying the minimum and maximum tracking point circles (green).
Enter the time in seconds for how long the active tracking points (red circle) should be displayed in the preview.
The damping allows reducing noisy input values. The damping factor is set by default to 0,5. This setting can be changed from 0 = maximum damping up to 1 = no damping.
As soon as the Face Tracker detects a face / body or eyes, it will be surrounded by a red circle.
The position data of this circle (x,y,z) is displayed top left of the Window.
To use this data for further processing, please refer to Facetracking Input Node.