|
Image metamorphosis (also commonly known as morphing) is described as "image warping with color interpolation. Image warping applies 2D geometric transformations on the images to retain geometric alignment between their features, while color interpolation blends their color." (Wolberg). Image morphing extracts feature primitives between two images which are then used to interpolate the positions of the features.
|
|
|
Mona Lisa and her reflection morph. (from http://www-2.cs.cmu.edu/~seitz/vmorph/vmorph.html)
|
|
There are many implementations of image morphing. Morphing algorithms are based on mesh warping, field morphing, radial basis functions, thin plate splines, energy minimization, and multilevel free-form deformations. In this project I will investigate feature-based image morphing and view morphing. Algorithms will be implemented in Matlab (OpenCV might be used for a final implementation) using images from the internet and images from a digital camera. The program will output N images that perform the morphing. These images can then be edited together into a short animated sequence.
|
|
The algorithms will be compared based on the resulting morph. We'll compare distortion in the image, any problems of ghosting, complexity of the algorithm, and time of execution. Although the analysis will be subjective, we hope to compare the strengths and weaknesses of the two algorithms. Using these results, we'll analyze the feasibility of creating a real-time automated version of view morphing, and attempt an implementation.
|
| Feb 4 | Project Proposal |
| Mar 7 | Algorithms Implemented |
| Mar 14 | Compare Algorithms |
| Mar 25 | Application to Real Time View Morphing |
| Mar 25 | Project Review |
| Apr 10 | Final Report |
|
|
|