|TORRES ROCIO NAHIME||Cycle: XXXIII |
Section: Computer Science and Engineering
Tutor: ALIPPI CESARE
Advisor: FRATERNALI PIERO Major Research topic
:Analysis of Geographic Data for Open Data sources enrichment and environmental monitoringAbstract:
environmental monitoring, hydrogeologic risk mapping, land use, and urban planning. However, most available data sets are obtained based on the human-based created entities. For example, open datasets are usually obtained through the spontaneous collaboration of users (this is known as Volunteer Geographic Information System - VGIS). This not only is a very time-consuming task, but can also result in the lack of completeness and, in some cases, of quality. To address this issue, a hybrid approach is proposed, in which geographical and anthropic entities (such as mountain summits, trekking paths, river beds, and other landforms) are automatically discovered by analyzing available sources (such as the NASA SRTM Earth Digital Elevation Model - DEM and public aerial images) and the confidence in the automatically extracted entities is improved by submitting (only) the objects with uncertainty over a threshold to a crowdsourcing step. The resulting objects can be provided to the crowd to be used as a baseline to map (through validation) new entities or to improve the information about the existing ones in a VGIS. Depending on the use case, the data analysis approaches may exploit heuristic algorithms, which typically depend on multiple, manually set, parameters and are hard to generalize to different Earth regions. In this research project, we aim to explore the use of Deep Learning methods to train a model capable of identifying geographical and anthropic entities with accuracy, at scale, and in a region-independent way, and at developing web application(s) to publish uncertain results within crowdsourcing campaigns for validation and later incorporation of the new objects into GIS datasets.