19 October 2016 Airborne hyperspectral sensor radiometric self-calibration using near-infrared properties of deep water and vegetation
Author Affiliations +
Abstract
Retrieving the water-leaving reflectance from airborne hyperspectral data implies to deal with three steps. Firstly, the radiance recorded by an airborne sensor comes from several sources: the real radiance of the object, the atmospheric scattering, sky and sun glint and the dark current of the sensor. Secondly, the dispersive element inside the sensor (usually a diffraction grating or a prism) could move during the flight, thus shifting the observed spectra on the wavelengths axis. Thirdly, to compute the reflectance, it is necessary to estimate, for each band, what value of irradiance corresponds to a 100% reflectance. We present here our calibration method, relying on the absorption features of the atmosphere and the near-infrared properties of common materials. By choosing proper flight height and flight lines angle, we can ignore atmospheric and sun glint contributions. Autocorrelation plots allow to identify and reduce the noise in our signals. Then, we compute a signal that represents the high frequencies of the spectrum, to localize the atmospheric absorption peaks (mainly the dioxygen peak around 760 nm). Matching these peaks removes the shift induced by the moving dispersive element. Finally, we use the signal collected over a Lambertian, unit-reflectance surface to estimate the ratio of the system's transmittances to its near-infrared transmittance. This transmittance is computed assuming an average 50% reflectance of the vegetation and nearly 0% for water in the near-infrared. Results show great correlation between the output spectra and ground measurements from a TriOS Ramses and the water-insight WISP-3.
Conference Presentation
© (2016) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Kévin Barbieux, Vincent Nouchi, Bertrand Merminod, "Airborne hyperspectral sensor radiometric self-calibration using near-infrared properties of deep water and vegetation", Proc. SPIE 9999, Remote Sensing of the Ocean, Sea Ice, Coastal Waters, and Large Water Regions 2016, 99990M (19 October 2016); doi: 10.1117/12.2241251; https://doi.org/10.1117/12.2241251
PROCEEDINGS
17 PAGES + PRESENTATION

SHARE
Back to Top