Design and validation of object recognition methodologies for underwater fluorescence LIDAR applications

Goal of this work is to present and validate an underwater object spectral recognition methodology for fluorescence LIDAR signals by using an underwater fluorescence LIDAR propagation model. The spectral recognition methodology is aimed at deciding if an underwater object detected in the water column can be identified as belonging to a data base of objects of interest characterized by known fluorescence spectral signatures. The methodology needs to compensate the received signal for the water column effects in order to derive an estimate of the underwater object fluorescence spectrum to be used for spectral recognition. By using an underwater fluorescence LIDAR propagation model developed ad hoc, the methodology may be validated in different system, geometric, and environmental conditions. Experimental results obtained in two different acquisition scenarios show that the underwater object recognition methodology is promising for recognizing objects submerged in the water column at different depths and highlight the utility of the developed LIDAR propagation model for assessing the object recognition performance that may be experienced in various different acquisition conditions.