This work demonstrates a new statistical approach towards backscatter “clutter” rejection for continuous-wave underwater lidar systems: independent component analysis. Independent component analysis is a statistical signal processing technique which can separate a return of interest from clutter in a statistical domain. After highlighting the statistical processing concepts, we demonstrate that underwater lidar target and backscatter returns have very different distributions, facilitating their separation in a statistical domain. Example profiles are provided showing the results of this separation, and ranging experiment results are presented. In the ranging experiment, performance is compared to a more conventional frequency-domain filtering approach. Target tracking is maintained to 14.5 attenuation lengths in the laboratory test tank environment, a 2.5 attenuation length improvement over the baseline.

Date Published: 17 May 2016
PDF: 10 pages
Proc. SPIE 9827, Ocean Sensing and Monitoring VIII, 98270J (17 May 2016); doi: 10.1117/12.2222817

Published in SPIE Proceedings Vol. 9827:
Ocean Sensing and Monitoring VIII
Weilin W. Hou; Robert A. Arnone, Editor(s)