Study On Beam Patten Correction Of Multisector Multibeam Sonar And Seabed Sediment Classification

With the development of acoustic detection technology,multibeam sonar,as a mainstream instrument for underwater terrain acquisition,using the recorded echo intensity data to investigate the distribution of seabed sediment has become an important application.At present,the multi-sector multibeam sonar that is commonly installed on China’s marine scientific research vessels has caused significant systematic beam pattern radiation distortion problems in the intensity data due to the working characteristics of multi-sector transmission,dual-swath acquisition,and survey mode switching,Severely limiting the value of data utilization.At the same time,the seabed sediment classification based on multi-beam sonar is mostly carried out using angular response curves or seabed backscatter mosaic images as data sources,which is affected respectively by low spatial resolution or nadir striped artifacts,resulting in inaccurate category boundaries and low classification accuracy or vacancies in near-nadir regions.Therefore,in this paper,the analysis of the radiation model of the multi-sector multi-beam system and the statistical characteristics of the seafloor backscatter is carried out,and the relative correction method of the multi-sector beam pattern and the classification method of the combination backscatter statistical characteristics and image structure information are proposed.The former solves the problem of systematic intensity distortion caused by the radiation characteristics of the multi-sector multi-beam system,and the latter achieves highresolution,full-coverage unsupervised seabed sediment classification.The main work and contributions of the paper are as follows:1.First,the acoustic principle of multibeam sonar backscattering intensity measurement,seabed sediment classification and the processing flow of echo intensity data of conventional single-sector system are introduced.Then,the defects of the traditional classification method using the angle response curve or the seabed backscatter image are analyzed.Finally,the characteristics of multi-sector multi-beam systems and their impact on backscattering intensity estimation are discussed.2.A relative correction method for multi-sector beam patterns is proposed.This method first selects a small amount of data to calculate the echo intensity of different seabed incidence angles according to the survey mode,acquisition swath,and transmission sector,and then uses the intensity estimated by the empirical backscatter angular model as a reference to determine the relative radiation correction value between sectors,and finally apply the correction value to all measurement lines to compensate for systematic radiation distortion problems.Experiments show that the method can effectively eliminate the systematic intensity distortion caused by the multi-sector multibeam sonar operation mode without using additional correction equipment or field experiments,and obtain high-quality seafloor backscatter intensity signals.Compared with before correction,the intensity curve between the vertical track and the adjacent sectors continuously changes without misalignment,and the intensity of the parallel track to different sectors at the same seabed incidence angle is basically consistent,improving the utilization value of the seabed echo intensity data of the multi-sector multibeam system.3.A classification method combining the statistical characteristics of seafloor backscatter signal and seabed image structure information is proposed.This method first constructs the probability density curve of each category based on the intensity histograms at different angles of the oblique incidence area,and secondly uses the inverse distance weighting method to interpolate the probability likelihood map of the full coverage of the measurement area,and then uses the guided filtering method to introduce image structural information optimizes the probability graphs of different categories,and finally uses Bayesian decision theory to obtain unsupervised classification results.Experiments show that the method combines the advantages of backscatter intensity signals and seabed mosaic images in sediment distinction and boundary determination to achieve high-resolution seabed sediment classification result within the full-coverage of the survey line.In the classification results,the boundary of the category is clear and the influence of speckle noise is small.Comparing the classification results of the Jiaozhou Bay with sediment sampling data shows that the number of automatically determined categories is basically consistent with the actual situation,and the overall classification accuracy rate can reach 86.3%,proving the correctness and rationality of the classification method.4.A large-area multi-beam seabed sediment classification application was carried out in a sea area of the South China Sea.First,the echo intensity data collected by the multisector multibeam sonar were corrected for beam pattern radiation distortion,and highquality backscatter intensity signals and seabed mosaic images were obtained.In the direction of the parallel track,the intensity difference caused by different sectors decreased from 1.98 d B to 0.31 d B.On this basis,a experiment was conducted to form a full-coverage,high-resolution unsupervised classification result in the survey area.The classification results were tested using cross survey lines,and the overall agreement rate reached 86.4%,further verifying the effectiveness of the beam pattern correction method and the superiority of the classification method.