Giovanni Marchisio
Director - Product Development and Labs
DigitalGlobe

Biography
Giovanni Marchisio is the Director of Product Development and Labs at DigitalGlobe, where he is developing scalable image exploitation and search solutions for the largest commercial satellite image library in the world. Previously he worked for thirteen years for Insightful/Mathsoft, the maker of the statistical data mining platform S-PLUS, most recently as CTO and Vice President of R&D. He was also a Principal Investigator on 14 R&D contracts with NASA, DARPA, the USAF, the US Army and the NIH, and led a pioneering project in image mining with NASA’s Goddard Space Flight Center. He has articulated novel ideas and architectures in the areas of multimedia retrieval, multispectral satellite image mining, and search engines.  He is the author of several papers, book chapters, the primary inventor on ten recent US patents, and was an invited speaker at the US House of Representatives in 2001. He has been a visiting professor with the Department of Geophysics and Astronomy at the University of British Columbia (UBC), Vancouver, Canada. He holds a B.A.Sc. in Engineering from UBC and a Ph.D. in Geophysics and Planetary Physics from the University of California San Diego (Scripps).

Abstract
The Future of Geospatial Big Data

DigitalGlobe's operates the most advanced constellation of satellites in the world. Their unparalled spatial and spectral resolution, accuracy, collection capacity, revisit times and large image archives provide the basis for creating a living digital inventory of everything on the surface of the planet. Through a combination of computer vision, machine learning, crowdsourcing and data fusion large volumes of raw imagery can be turned into actionable information. The extraction of novel geospatial big data layers from sub meter imagery at scale unlocks unprecedented opportunities for information services by supplementing more conventional data sources and offering inexpensive refresh over large areas. Examples of these information layers include 2m land use land cover maps, automated crop identification and inventorying, monitoring of human presence and activity, detection of man-made structures at sub-meter scales and high resolution population density estimates in continents like Africa where over 70% of the population is not located, detection of oil tanks and measurement of oil reserves, counting objects like cars and trucks in the entire country of Japan, automatically detecting new constructions and building improvementsover the entire Moscow district, counting all the containers in the harbor of Vancouver and measuring car factory output on a country scale. Information layers not only show change over time, but can actually help predict future human and environmental activities orbe quickly be generated and deployed in support of natural disaster and crisis event management. In cooperation with partners worldwide, we are developing insight solutions that leverage these unique geospatial big data layersto provide a dynamic HD view of our globe.