Page tree

Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

The mentioned camera parameters (1920 x 1080, 30 fps, mono) should be appropriate for our software and also for most of the use cases.

Warning

This text requires review after the introduction of new face detector in V8.7b2. Also, https://visagetechnologies.com/facetrack-features/ needs to be reviewed.

In general, visage|SDK works with a very wide range of camera resolutions and can provide sustainable tracking on faces as small as 30×30 pixels. For further details on inputs and other features specifically for face tracking please see https://visagetechnologies.com/facetrack-features/

...

How far from the camera can the face be tracked/detected?

Warning

This text requires review after the introduction of new face detector in V8.7b2.

The tracking distance depends on the camera resolution. For the face to be detected and tracked, it should be at least 30 pixels wide in the image. For example, for a webcam with resolution 1920×1080 (Logitech C920), tracker can be configured to detect and track faces up to ~7.25 meters from the camera (with performance tradeoff).

...

This depends on camera resolution. Face Recognition works best when the size of the face in the image is 150 at least 100 pixels or more.

Info

See also:

...

  • Face tracking is limited to 20 faces (for performance reasons). To locate more faces in the image, use face detection (class FacialFeaturesDetector).There is a performance trade-off in face detection: detecting smaller faces

  • visage|SDK detects/tracks faces whose size in the image is computationally more expensive. at least 5% of the image width (height in case of portrait images).

    • If you are using high resolution images with many faces, so that each face is smaller than 5% of image width, one solution may be to divide the image into portions and process each portion separately; alternatively, a custom version of visage|SDK may be discussed.

  • For the face to be detected, it should be about 30 pixels wide in the image. For face recognition and analysis (age, gender, emotion), higher resolution is better, ideally 100-150 faces should ideally be about 100 pixels wide.

Troubleshooting

...