Abstract
Backscatter mosaicking is a necessary step in the analysis and interpretation of sidescan and multibeam sonar records. However, due to limitations intrinsic to the mosaicking technique, backscatter mosaics are restricted in their capacity to unambiguously discriminate seafloor properties. A more adequate technique to characterize the seafloor is the analysis of backscatter angular responses, since those responses are intrinsic properties of the seafloor. This technique sometimes lacks spatial resolution, however, as the analysis is limited to the swath width of the sonar. In this paper, we propose an approach to combine mosaicking and angular response analysis techniques in an attempt to take advantage of both the spatial resolution of the mosaic, and the angular resolution derived from the angular response analysis. In order to test these ideas, we used acoustic backscatter acquired by a Reson 8101 (240kHz) multibeam sonar during normal survey operations conducted on the NOAA Ship FAIRWEATHER around Cape Decision, Alaska in spring 2005. First, we defined parameters that uniquely described the angular responses, and treated those parameters as a feature vector in a multidimensional space. The parameters were then clustered with a simple unsupervised clustering algorithm. The result of the clustering analysis defined areas on the seafloor which had similar angular responses, which we called themes. We then used these themes to develop more robust indicators of angular response from their coverage areas, which were finally used as Angle Varying Gain correction tables to assemble an enhanced mosaic.
Department
Center for Coastal and Ocean Mapping
Publication Date
5-2007
Journal Title
U.S. Hydrographic Conference
Conference Date
May 14 - May 18, 2007
Publisher Place
Norfolk, VA, USA
Publisher
Hydrographic Society of America
Document Type
Conference Proceeding
Recommended Citation
Fonseca, Luciano E. and Calder, Brian R., "Clustering Acoustic Backscatter in the Angular Response Space" (2007). U.S. Hydrographic Conference. 384.
https://scholars.unh.edu/ccom/384