Visible and infrared imaging spectroscopy for high resolution mapping and health assessment of Maine’s forest and agricultural resources
PI: Nelson, Peter (Biological and Environmental Sciences, UMaine)
Sector: Forestry, Information Technology
Partners: UMaine
Abstract: Our goal is to integrate ground-based spectral scanning/chemical analysis and data mining of hyperspectral images into a pipeline for detection of specific, user-generated targets (e.g. specific plants, pathogens, stress signals, etc..) for Maine’s economically important natural resource sectors and elsewhere for competitive research applications. A graduate student would improve this hyperspectral image processing capacity using existing imagery synergized with our current spectroradiometric and UAV-image acquisition capacity. The image processing would focus target sites with existing imagery from NASA contacts connected to key economically important crops, specifically forest resources flown by G-LIHT last year. The student would help develop new and better algorithms for mapping, utilizing to the very sensitive cameras and co-located additional datasets with high resolution reference data, which enables detecting problems (e.g. insects or disease) or positive signals (exceptional growth) in which managers could then act to either mitigate disease or stress or expand certain treatments found to be exceptionally beneficial.