Invited Talk - Dr. Raju Vatsavai
Ontology and Semantics in Very High-Resolution Image Information Mining
Image classification can be of two types: thematic classification and semantic classification. In thematic classification, thematic class labels (e.g., buildings, roads, crops, forest) are assigned to pixels, using some discriminant function. There are several challenges in remote sensing image classification, some due to the sensor capabilities (spectral, spatial, and temporal resolution) and others are due to the definition of thematic classes. Though spectral resolution dictates how fine-grained classes one can obtain from the classification, many times thematic classes are often aggregated (e.g., hardwood, conifer other forest types are aggregated into simply “forest”). However, we show that ontologies can be used to derive fine-grained classification from limited ground-truth and aggregated classes. On the other hand, very high-resolution images allows us to identify semantic classes such as nuclear power plants and slums, and not just thematic classes such as buildings. In this talk, we present recent advances in image information mining where ontologies and semantics allow one to extract fine-resolution and semantic classes from remote sensing images.
Biography - Dr. Raju Vatsavai
Dr. Raju is a Chancellor’s Faculty Excellence Program Cluster Associate Professor in Geospatial Analytics in the Department of Computer Science, North Carolina State University (NCSU). He works at the intersection of spatial and temporal big data management, analytics, and high performance computing with applications in the national security, geospatial intelligence, natural resources, climate change, location-based services, and human terrain mapping. As the Associate Director of the Center for Geospatial Analytics (CGA), Raju plays a leadership role in the center’s strategic vision for spatial computing research. Before joining NCSU, Raju was the Lead Data Scientist for the Computational Sciences and Engineering Division (CSED) at the Oak Ridge National Laboratory (ORNL). Raju has more than 25 years of research and development experience in large-scale spatiotemporal data management and geographic knowledge discovery.