Research On Abnormal Correction Of Central Beam Data Of Multi-beam Sonar And Sonar Image Processing Technology

The multi-beam sonar system can obtain high-precision maps of the seabed topography and geomorphology through sounding data and backscattering intensity data,which can provide important information for the development of marine resources.However,the current multi-beam backscattering intensity data processing and sonar image processing still have deficiencies.To this end,based on the principle of the multi-beam sonar system,based on the analysis of related research at home and abroad,the following research has been carried out :(1)The composition structure and working principle of the multi-beam sonar system are introduced in detail,the process principle and influencing factors of the multi-beam sonar image formation are analyzed,and the processing flow of the multi-beam backscattering intensity data is given,including : TVG removal,sound wave propagation loss correction,sound illumination area correction,incident angle effect correction,etc.during acquisition.Aiming at the phenomenon of missing data in some areas of the multi-beam sonar image,three methods were studied to correct it.The bilinear interpolation method,inverse distance weighted interpolation method,and kriging interpolation method were used to correct the missing areas of the sonar image.Experiments show that the kriging interpolation method averages the error of the sonar image in the process of processing the missing areas.Smaller and better.(2)Analyze the cause of the central beam anomaly in the multi-beam sonar image,introduce the general processing method of the central beam anomaly,and propose a method for correcting the central beam region anomaly based on wavelet analysis.This method uses the multi-resolution features of wavelet transform to extract the high-frequency signals in the central beam region and the low-frequency signals on both sides of the hull respectively.The trend line of the corrected central beam region is calculated using the overall least squares method,and it is viewed after the correction.The trend line of is the discrete smooth component,and the high-frequency signal in the central beam region is the detail component.Finally,the backscattering intensity of the central beam region is reconstructed according to the signal recombination characteristics of the wavelet transform.Experiments show that this method can effectively correct the abnormal phenomenon of the multi-beam central beam region,and has a smaller standard deviation and better continuity than the sonar images obtained by other correction methods.(3)The multi-beam sonar image mosaic processing flow is introduced,and the ghosting problem caused by the traditional sonar mosaic method is analyzed.By studying the mosaic of multi-beam sonar images based on feature matching,the SURF algorithm is used to mosaic the multi-beam sonar images of the same band and adjacent bands.Experiments show that the nato image mosaic method based on the SURF algorithm effectively solves the ghosting problem of the traditional mosaic method,and greatly improves the mosaic accuracy of the multi-beam sonar image.