Post-processing Methods Of Multibeam Echo Sounder Backscatter Strength Data And Sonar Image Mosaic

Obtaining detailed seabed topographic and geomorphological information is fundamental for exploring natural resources under the seas.Multibeam Echo Sounder(MBES)is just such a tool for acquiring underwater high-density bathymetric data and the backscatter strength(BS)information of relative seafloor sediment.The accurate high-resolution images of seabed topography and geomorphology can be obtained through rigorous processing of the bathymetric and backscatter strength data.However,it still remains as a challenge task to accurately process backscatter strength data and mosaic sonar images.Developing new processing methods for backscatter strength data and sonar image mosaic is focused in this study.As a result,a couple of new methods are developed which enable to obtain more reliable and accurate seabed topographic and geomorphic images comparing to some existing methods.This study covers following topics:(1)The current state of MBES data processing methods of angle response correction of backscatter strength data and the MBES sonar images mosaic are reviewed.Based on analysis of current methods,the objectives of this study were outlined.(2)Various analysis is performed for multibeam data processing,which includes the data decoding,sound ray tracking,spatial location,sampling point extraction,backscatter strength compensation,and sonar image preprocessing.A set of new methods are then developed.A high-resolution sonar image,which is only affected by angular response,could be obtained using the proposed methods.(3)The angular response has severe impacts on backscatter strength data,current correction models didn't address this issue in great extents.Two angular response correction methods are proposed: a)improved angular response correction method which takes high-incidence region into account;and b)Gauss fitting correction method.The former one is based on the actual acoustic wave scattering principle,and the latter one is based on the backscatter strength change regulation.Using either angular response correction methods could reduce the impacts of angular response in great deal.(4)A SURF(Speeded Up Robust Features) based sonar image mosaic method is developed which minimizes the residual differences among overlapped sonar images.A strategy of using coordinate constraints during feature matching is employed,which effectively eliminates the majority of outliers and therefore improves the matching accuracy.This method is successfully applied to fuse multiple sonar images to form a seamless mosaic.