SBIR Phase II:High-Resolution Image Segmentation for Natural Resource Management

The commercial/broader impact of this Small Business Innovation Research (SBIR) Phase II project is to provide economic benefits, health advantages, and improved natural disaster response readiness to the USA. Historically, natural resource and conservation organizations have had difficulties mapping their targeted ecosystems, whether due to high costs of manual surveys or poor resolution of imaging technologies. Annually, organizations spend more than $27 billion on geospatial monitoring and analysis. This Phase II project will decrease the cost of ecosystem mapping while increasing resolution, allowing for the best quality vegetation health tracking available. Additionally, this project will result in a 50-90% reduction in work hours for natural resource mapping. By saving time, stakeholders can allocate effort to other aspects of natural resource management. By mapping land use over time, managers and conservationists can track land changes and determine if currently-implemented programs are having intended impacts on the ecosystem. This project will also improve monitoring and managing of vegetation across watersheds that provide roughly 80% of US drinking water - systems where water quality relies on healthy and biodiverse vegetation to filter pollutants. Lastly, this project will improve the ability of government agencies to rapidly monitor environmental impacts of natural disasters and inform responses._x000D_ _x000D_ This Small Business Innovation Research Phase II project will develop a comprehensive software system that can provide unparalleled spectral and spatial detail on diverse landscape scenes. Compared to current labor-intensive field testing, this project’s outputs will offer scene characterization at comparable, or better, levels of detail, while surmounting the time, cost, and accessibility constraints that have historically precluded comprehensive and repetitive monitoring. Accomplishment of these Phase II goals will yield a user-friendly land cover mapping system that will enable high-resolution environmental monitoring. System outputs on population dynamics, climate change-induced vegetation shifts, and disease assessments can facilitate data-driven decision-making for precision ecosystem management and climate action. The framework of the innovation consists of three main components: 1) image pre-processing and alteration, 2) image segmentation, and 3) resolution recovery. This approach provides rapid replicability between ecosystem types and versatile scalability due to processing efficiency, while providing currently unavailable ecosystem health indicators._x000D_ _x000D_ This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.