Automatic Imagery Ortho-Rectification Pilot
Project Leads and Affiliations:
E. Terrence Slonecker, EPA/Office of Research and Development
Carolyn Offutt, EPA/Office of Solid Waste and Emergency Response
Chris Morrell, US Central Intelligence Agency
Project Abstract Written: October 2005
Project Timeframe: 2006 - 2008
Remote Sensing is becoming an increasingly important science for advancing understanding of environmental processes, conditions, and changes, and for both human and ecological health. Key societal decision-making processes routinely rely on information derived from imagery and remote sensing systems. However, even with the significant advancements in sensor technology, one of the most important and enduring and important aspects of remotely sensed data is the simple fact that each image represents a sample point in time that is preserved for countless future applications. Studies that utilize historical imagery, such as change detection, are among the most important uses of remote sensing data. To realize the full breadth of remote sensing information potential relevant to environmental decision-making, the enormous archives of analog remote sensing data must be converted to digital format and geo-referenced into a geodetic coordinate system.
This research effort will develop and test an automated, ortho-rectification process (AORP) that would enable the timely and cost-efficient geo-registration of analog and/or raw aerial photography and other remote sensing data and the automated development of digital metadata that would in turn facilitate internet-service based distribution of these data. Currently, the enormous archives of historical aerial photographs must be manually digitized and ortho-rectified, which is time-consuming and labor-intensive, and generally preludes the routine use of these data in many routine spatial applications. An automated geo-registration process would enable the practical development and application of digital geo-registered historical imagery databases that could help realize the full potential of historical imagery for GEOSS and other environmental applications.
This project will develop an AORP system in four steps: First will be the development and testing of both a specific aerial photo, and a generic senor model, based on previous successful applications with commercial satellite systems. Second, three parallel Regional demonstration projects will perform automated ortho-rectification on existing analog holdings of historical aerial photos and/or satellite imagery, assess the efficiency in terms of cost and labor-hour savings, and deliver that digital information to the appropriate EPA region office for use in environmental programs.The full utilization of historical aerial photos and other remotely sensed data, dating over 70 years, is often problematic due to the analog nature of these holdings. Manual analog to digital conversion is time-consuming and costly and precludes, from a practical standpoint, the full-scale utilization of these data. This project represents an opportunity to significantly increase the base of remotely sensed information and to efficiently serve that information, via GIS/internet services, directly to the EPA and other GEOSS decision-makers.