Merging VRML Models: Extending the Use of Photomodeller

Hudson, Thomas Randall, Department of Computer Science, University of Virginia
Martin, Worthy, Department of Computer Science, University of Virginia

This project resulted in the initial development of a program that helps in the creation of complex three-dimensional models. The program's original purpose was to extend the use of the commercially available package PhotoModeler. The program will first be used by the Ancient Reconstruction of Pompeii project at University of Virginia's School of Architecture. The program extends PhotoModeler by stitching together smaller models created by PhotoModeler. These models use the Virtual Reality Modeling Language (VRML) format. The program was developed with PhotoModeler in mind; however, it should also work with any source of VRML files. The stitching process relies on overlapping regions among pairs of models. After some necessary background information has been presented, this report describes the algorithms used to find the similarities between models and to align the models based on these similarities. When aligning models, the program works within three degrees of freedom: scale, rotation, and translation. Initial testing on a small sample shows promising results. The program passes all available test cases. The program is far from finished, but it correctly aligns user-specified pairs of models. Future work on the program will incl ude stitching a pair of models and writing the resulting model to disk. Efforts will also be made to further automate the program. The desired finished product is one that will merge an entire set of models with minimal user intervention.

BS (Bachelor of Science)

Thesis originally deposited on 2011-12-28 in version 1.28 of Libra. This thesis was migrated to Libra2 on 2016-11-30 15:15:21.

All rights reserved (no additional license for public reuse)
Issued Date: