Interpolating Pixel Art
| dc.contributor.advisor | Gingold, Yotam | |
| dc.contributor.author | Rutherford, Anson | |
| dc.creator | Rutherford, Anson | |
| dc.date | 2018-12-07 | |
| dc.date.accessioned | 2019-07-01T19:24:19Z | |
| dc.date.available | 2019-07-01T19:24:19Z | |
| dc.description.abstract | We present a method of generating plausible interpolations between frames from ani- mated pixel art. We do this in the hopes of helping artists working in the style by providing a simple method of interpolation, which would reduce the amount of by-hand work required by the artists. We adapt an existing method for image interpolation, which uses energy cost minimization to define motion for each pixel position in the image. The specific challenges inherent in interpolating pixel art are discussed and examined. We describe the reasons the original paper's methods won't adequately work for our specific problem. Then, we describe the changes we make to the original paper's energy cost function to attempt to accommodate those challenges provided by pixel art. These changes focus on gathering more information from the source images, to make up for the limited amount of color infor- mation. We examine these changes and their results by interpolating the same images with different weights for the various factors we propose. We discuss which methods work best in general, and in specific cases. | |
| dc.identifier.uri | https://hdl.handle.net/1920/11466 | |
| dc.language.iso | en | |
| dc.subject | Interpolation | |
| dc.subject | Pixel art | |
| dc.subject | Animation | |
| dc.subject | Optical flow | |
| dc.title | Interpolating Pixel Art | |
| dc.type | Thesis | |
| thesis.degree.discipline | Computer Science | |
| thesis.degree.grantor | George Mason University | |
| thesis.degree.level | Master's | |
| thesis.degree.name | Master of Science in Computer Science |