Mighty Morphs
- Image Morphing
Project 3

The idea of averages is a powerful tool when looking at a large set of data. If you are looking at test scores, to sports statistics, to gambling, to machine learning, to signals and systems: averages provide the general feeling for a given situation. This idea can be applied to a set of images as well. If the average of a set of images is computed, the outcome is rather unique (or I should say strikingly not unique). The average image will show what is consistent between all of the images, it will show the general feeling of all of base pictures.

This project focuses on images of the classes' faces. The first part of the program computes the blend from one face to another. It is important to note, that this is not just a cross-dissolve, that would give you an image that could be considered an "average," but it is not as clear as if you average the geometry of the faces, and then dissolve. The second part of the project deals with all of the faces of the class. An average geometry is computed, every face is warped in to that geometry, and then the results are blended together. It is hard to exactly describe the outcome... go look at the galleries! Be amazed!

For further reading:
The Original Project Description.