Fusion Through Interpretation
Abstract
We discuss two problems in the context of building environment models from multiple range images. The first problem is how to find the correspondences between surfaces viewed in images and surfaces stored in the environment model. The second problem is how to fuse descriptions of different parts of the same surface patch. One conclusion quickly reached is that in order to solve the image-model correspondence problem in a reasonable time the environment model must be divided into parts.
Cite
Text
Orr et al. "Fusion Through Interpretation." European Conference on Computer Vision, 1992. doi:10.1007/3-540-55426-2_91Markdown
[Orr et al. "Fusion Through Interpretation." European Conference on Computer Vision, 1992.](https://mlanthology.org/eccv/1992/orr1992eccv-fusion/) doi:10.1007/3-540-55426-2_91BibTeX
@inproceedings{orr1992eccv-fusion,
title = {{Fusion Through Interpretation}},
author = {Orr, Mark J. L. and Hallam, John and Fisher, Robert B.},
booktitle = {European Conference on Computer Vision},
year = {1992},
pages = {801-805},
doi = {10.1007/3-540-55426-2_91},
url = {https://mlanthology.org/eccv/1992/orr1992eccv-fusion/}
}