Skip to content

Image Warping#

Warping is the process of face image normalization. It requires landmarks and face detection (see chapter "Detection facility") to operate. The purpose of the process is to:

  • compensate image plane rotation (roll angle);
  • center the image using eye positions;
  • properly crop the image.

This way all warped images look the same and one can tell that, e.g., left eye is always in a box, defined by the certain coordinates. This way certain transform invariance is achieved for input data so various algorithms can perform better.

Face warping
Face warping

Be aware that image warping is not thread-safe, so you have to create a warper object per worker thread.

Back to top