Dr. Craig Knoblock - With the explosive growth of the Web combined with the widespread availability of satellite imagery through systems such as GoogleEarth,there are now a huge number of geospatial data sources available. An important challenge with the availability of the goespatial data is how to locate, extract, and integrate the data to fully exploit the information. In this talk I will describe our recent work on finding, extracting and aligning data from online raster maps. This includes machine learning techniques to automatically classify images to find maps, pattern recognition techniques to automatically extract road building, and text information from maps, and geospatial integration techniques to accurate fuse maps and their extracted layers with high-resolution orthoimagery. Additional Public Info: Reception: 3:30-4:00 348 Avery Hall Talk: 4:00-5:00 PM 115 Avery Hall