Fusion of hyperspectral imagery and bathymetry information for inversion of bioptical models

Bioptical models are used jointly with hyperspectral imaging in inversion procedures for mapping of benthic habitats. Several algorithms have been described in the literature to remove the effects of the water column and extract information about the sea bottom that only take into consideration the measured hyperspectral image. However the availability of LIDAR derived bathymetry information opens the possibility of using this information for improved retrieval of the bottom properties. We present in this paper a study using simulated and hyperspectral imagery on the improvement in benthic habitat mapping that can be achieved by fusing bathymetry and hyperspectral imagery. Simulation results show that it is possible to obtain accurate bottom abundance estimates 5-10 meters beyond what can be obtained with hyperspectral imaging alone in clear waters. With real data we demonstrate increase in accuracy with respect to ground truth.