One problem with the feature-based morphing technique of Beier and Neely is that all features had to be specified by the animator. However, in highly correlated images, such as faces, finding features should be able to be found automatically (or semi-automatically, which I will discuss below). For instance, the view morphing technique of Seitz and Dyer can be used to automatically align faces. Once a face is aligned, salient features can be found using template matching, or perhaps some other technique.
I hope to be able to find these salient features of faces. I expect to use a lot of high-level (i.e. trial and error) information to figure out which features are most important and how those features should be specified. I expect to provide at the least a very rough estimate of the control points guiding the morph for the user to correct by hand, but I would like to automate the system as much as possible.
| Date | Activity |
| 4/6 - 4-11 | Background Reading |
| 4/12 - 4-18 | Implement Beier and Neely |
| 4/19 - 4-25 | Implement Seitz and Dyer for face alignment |
| 4/26 - 5-4 | Finding Features |