Context Aware Object Geotagging
This work addresses geotagging accuracy for street objects like traffic lights, which is incremental as it builds on existing methods with contextual refinements.
The paper tackles the problem of improving object geolocation from street view imagery by enhancing image metadata using Structure from Motion and refining predictions with contextual geographic information from OpenStreetMap, achieving validation against state-of-the-art approaches for traffic light geotagging.
Localization of street objects from images has gained a lot of attention in recent years. We propose an approach to improve asset geolocation from street view imagery by enhancing the quality of the metadata associated with the images using Structure from Motion. The predicted object geolocation is further refined by imposing contextual geographic information extracted from OpenStreetMap. Our pipeline is validated experimentally against the state of the art approaches for geotagging traffic lights.