Abstract
To address imprecisely placed entrance markers for geospatial entities in digital maps, for instance, for new or low-traffic locations with sparse data, a multi-agent computational system can be utilized. The system can process various data sources, including user-supplied media such as photos and videos, to generate a three-dimensional (3D) model of an entity's structure and identify potential entrances. Concurrently, the system may construct a 3D contextual model of the surrounding region using data such as building footprints and road networks. An iterative alignment process can position the entity's 3D model within the regional model based on a set of logical rules, such as orienting an entrance toward a street. Once a suitable alignment is achieved, the system can project the entrance's location into a real-world geodetic coordinate. This process may improve the positional accuracy of entrance markers for navigational and mapping applications.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Melvix, Lenord, "Estimating Geospatial Entrances via Multi-Agent 3D Contextual Alignment", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/9202