Thursday, August 7, 2008 - 10:30 AM

OOS 19-8: Linking ecosystem and resource mapping to applications of automated feature extraction

Ian Hanou, NCDC Imaging

Background/Question/Methods

Protecting natural resources and preserving the highly individual ecosystems in the world today is fast becoming a high priority for all peoples and countries as it is widely acknowledged now that human growth and encroachment have negative impacts on the environment.  Impacts as varied as air quality and contamination to species extinction and lack of diversity have us all looking for innovative ways to take control of protecting and managing our lands and reducing environmental ruin.

Understanding what is needed, begins with understanding what you have. Interdisciplinary professionals are improving upon and developing techniques in automated feature extraction, for just that reason. Utilizing high resolution satellite imagery and aerial photography to produce land use and land cover maps for a myriad of clients who defined a need to inventory and manage vital resources.

Mapping vegetative and non- vegetative land cover in urban, agricultural, and natural settings for use in more effectively monitoring and managing land and resource usage has been the focus of many projects recently concluded and upcoming at NCDC. The analyses provided yield not only a means for monitoring the usage but also for evaluating the economic and aesthetic benefits of protecting ecosystems.

Automated feature extraction is a relatively new way of classifying land cover from high-resolution imagery and to delineate areas of interest that are mutually exclusive. To reduce processing time and increase accuracy, NCDC has developed a proprietary analytical process using algorithms which maximize the power of an object oriented machine learning approach to automated feature extraction.

Results/Conclusions

NCDC performed an Urban Tree Canopy analysis for Providence and quickly and accurately gave the City a powerful tool to help officials, partners, and the public understand what the existing canopy consisted of, how the city can benefit both economically and socially, and exactly what the potential for improved canopy was.

In another study, NCDC developed land cover mappings for agricultural co-ops in South America to help the agencies establish boundaries and to visualize crop rotations through change detection for better use of the land and to capitalize on economic improvements.

Participants will leave this presentation with a better understanding of applications and uses of AFE technology and its proven uses in land cover mapping for natural resource and ecosystem management, urban tree canopy analysis, and various other ecological issues. Data requirements, project components, timelines and cost examples to perform these types of analyses will also be provided.