Online face morphing

Morphing - computerized creation of an animation in which one image gradually turns into another.

Morphing processes explained

Morphing processes involve digital signal processing algorithms in order to analyze two images, and gradually and continuously transform one image into another.

An explanation of the morphing process can be found in the following video:

Face morphing

Face morphing is probably the most widely spread use of morphing algorithms. It has many applications:

  • Animation - morphing can be used in order to animate faces, in example produce such effects as smiling.
  • Visualizing aging process - set the start image to a certain age, end image to a certain older age of the same person, and morphing between the two would allow visualization of the aging process of the person.
  • Morphing between two different people - set start and end images as pictures of two different people, and you get to visualize continuous transformation between the faces of the two people.

Face morphing applications

As face morphing algorithms become more available to the public, applications that utilize face morphing appear online - free to end users. Some of them are covered below. - probably the most prominent online morphing tool. It allows averaging (blending) between two chosen images. The power of this tool is behind the pre-processing that is required. The user uploads two images of faces, and then goes through a predefined process of providing the morphing engine with locations of specific face parts on the image. This allows for a great morph quality compared to other tools out there. - this set of tools was developed by the University of St Andrews in UK. It has the following features:

  • Changing the age, sex and ethnicity of a person on an image
  • Face averager - make a static morph between two predefined images
  • Real time face morphing - provides a series of continuous images, all the way from start image to end image

The biggest disadvantage of these tools is the morph quality (compared to for example). This is due to the fact that the morphing algorithm automates the detection of face parts (eyes, nose, mouth etc), and evidently the outcome of such automation is poor compared to human input. - provide a face morphing application similar to The biggest difference being that it's software available for download (and not an online tool). It offers automated spots detection, as well as user defined detection (to bring better results).

A somewhat funnier (and less useful) tool was launched by Burger King as an online promotion technique -
This tool allows uploading an image, and turning it into a Simpson character! After upload, you feed the algorithm with basic data as the age, color, sex, and other facial features of the person on the image. Nice promotional trick at the least.

copyright ©