Riverine feature extraction from high-resolution imagery
Physical infrastructure designed and developed to enable this growth has been in place for 50 years or more, and an increasing proportion has reached the end of its design life. Taking dams for example, while the Canadian case is less well documented, 85% of U.S. dams will be >50 years old by 2020. In particular, decayed riverine infrastructure (e.g., dams, levees, bridge and pipeline crossings, etc.) has significant implications for public safety, economic and environmental health. However, identification, assessment and inventory of riverine infrastructure using traditional field surveys are often logistically difficult, expensive and slow to respond to emergency. This proposal is targeted to develop innovative image-processing algorithm that extracts riverine features (natural and infrastructure-related). The algorithm is science-based, accurate and can be rapidly deployed on a local or regional scale, which is urgently required by the partner company applying to projects and contracts with local and regional organizations.