GIS applications now increasingly make use of geo-located multimedia data such as images and videos. Furthermore, the wide-spread availablity of smartphones allows the acquisition of user-generated videos that are annotated with geo-properties. The sensor meta-data, e.g., GPS and digital compass values, are considerably smaller in size than the visual content and are helpful in effectively and efficiently manage and search through large repositories of videos. However, a major practical issue is the noisy nature of such sensor data. For example, due to sensor data inaccuracies the visual coverage described by the meta-data may not exactly match the actual video scene, which leads to imprecise search results and positional disagreements on map overlays. Obstructions between the camera and its captured objects make these situations worse. Therefore, robust error-tolerance is an essential feature of any geo-tagged video search application.
To this end we introduce HUGVid, a modeling and indexing approach for uncertain geo-tagged videos. We construct an uncertainty model for video frames and segments. Since the frame-by-frame uncertainty model involves high computational complexity, we then propose an approximate modeling method based on a video segmentation algorithm which eliminates costly overlap calculations between the query region and individual frames. Finally, we test the performance of HUGVid with both two real-world and a large-scale synthetic dataset. Experimental results show that our method achieves high precision and good scalability and allows the efficient querying of noisy sensor data. HUGVid also returns confidence probabilities with the results which can then be beneficially used in upstream GIS applications.
(Publisher abstract provided.)