Thesis
Methods for creating corner colored Wang tiles
Washington State University
Master of Science (MS), Washington State University
2010
Handle:
https://hdl.handle.net/2376/101040
Abstract
Corner and edge colored Wang tiles are used in computer graphics to synthesize large seamless and non-periodic textures while reducing the storage requirements that expansive textures require. Wang tiles have matching constraints that dictate which tiles may be placed next to each other to ensure a seamless transition between adjacent tiles; while using a finite set of tiles for use in tiling limits the storage required to generate texture. The most popular technique in the current literature for constructing Wang tiles is to combine samples from an image and perform a graph cut between four abutted samples and an overlay sample where the graph cut should find the cut that selects pixels that creates the least objectionable tile. This thesis introduces new methods for generating corner colored Wang tiles and discusses the important characteristics in generating them. Synthesized Wang Tiles (SWT) use a controlled texture synthesis algorithm for tile construction. Bilinear Blended Wang Tiles (BLWTiles) use blending for tile construction and Gauss Laplace Pyramid Blended Wang Tiles (GLWTiles) use blending over different levels of Laplacian pyramids for tile construction. BLWTiles and GLWTiles are evaluated procedurally and can be mapped to GPU evaluation, while the SWTile method is intended to precompute a set of Wang tiles. The BLWTile, GLWTile, and SWTile techniques can supplement existing methods for creating Wang tiles. The characteristics that influence the quality of corner colored Wang tiles can serve to improve new and existing tile creation techniques.
Metrics
9 File views/ downloads
29 Record Views
Details
- Title
- Methods for creating corner colored Wang tiles
- Creators
- Michael Joseph Persons
- Contributors
- Wayne O. Cochran (Degree Supervisor)
- Awarding Institution
- Washington State University
- Academic Unit
- Electrical Engineering and Computer Science, School of
- Theses and Dissertations
- Master of Science (MS), Washington State University
- Publisher
- Washington State University; Pullman, Wash. :
- Identifiers
- 99900525381901842
- Language
- English
- Resource Type
- Thesis