Testing different Methods of Face Tracking using OpenCV





Rating is available when the video has been rented.
This feature is not available right now. Please try again later.
Published on Mar 11, 2013

This video shows two different Face Tracking implemenations. First a Haar Cascade using the Viola-Jones method and some flipping to detect profiles. This method is limited by the orientation of the head (you can see the track get lost when the head is tilted to the side. The second implentation uses a Haar Cascade to initially find the face and them switches to tracking features frame by frame using Lucas-Kanada Optical Flow. Here, there is no head tilt limitation. All algorithms are on straight RGB data. We will be adding a depth stream which will make it even more robust. The last part of the sequence shows a first pass at knocking at the background using very noisy depth data.

  • Category

  • License

    • Standard YouTube License


When autoplay is enabled, a suggested video will automatically play next.

Up Next

to add this to Watch Later

Add to

Loading playlists...