Requirement Study on a Specialized Hyperspectral Aerial Imaging System for Marine Plastic Litter Classification

This paper presents the predesign research for an affordable hyperspectral aerial measurement system for short wave infrared wavelengths. It is shown which design specifications need to be considered and a solution approach for aircraft related tradeoffs is presented, as well as an assessment for data quality and system design parameters for material classification for maritime plastic litter. It is presented that an affordable approach for broad aerial monitoring methods with hyperspectral technology can succeed from a technical point of view. To quantify required data quality, laboratory generated data are used, artificially impaired and analyzed in terms of acceptable noise impairment. Findings as to how high the noise impairment can be for expected classification results of the chosen plastic litter materials are given.