Multi-beam And Side Scan Sonar High-quality Measurement Information Acquisition And Superposition

The multibeam echo sounder(MBES)has been widely applied in the marine surveying for providing high-resolution seabed topography.While the side-scan sonar(SSS)is the most widely used tool of obtaining the seabed image.A MBES can simultaneously obtain the bathymetric terrain and the seabed image.But the low resolution of the sonar image is often unable to meet the demand.The SSS can provide high-precision seabed images but with low location accuracy.In order to obtain the high-precision and high-resolution seabed topographic and sediment images by superimposing the SSS image on the MBES bathymetric terrain,this paper has carried out research and analysis in terms of MBES sounding data,MBES images,SSS images,and superimposing the SSS image on the MBES bathymetric terrain.1)Weaking the residual errors in the outer beams of MBES data based on the frequency spectrum characteristics of the seabed terrainHowever,some factors degrade the precision of bathymetry,including the sound velocity,the vessel attitude,the misalignment angle of the transducer and so on.Although these factors have been corrected strictly in bathymetric data processing,the final bathymetric result is still affected by their residual errors.In deep water,the result usually cannot meet the requirements of high-precision seabed topography.The combined effect of these residual errors is systematic,and it's difficult to separate and weaken the effect using traditional single-error correction methods.Therefore,the paper puts forward a new method for weakening the effect of residual errors based on the frequency-spectrum characteristics of seabed topography and multibeam bathymetric data.Four steps,namely the separation of the low-frequency and the high-frequency part of bathymetric data,the reconstruction of the trend of actual seabed topography,the merging of the actual trend and the extracted micro-topography,and the accuracy evaluation,are involved in the method.Experiment results have proved that the proposed method could weaken the combined effect of residual errors on multibeam bathymetric data and efficiently improve the accuracy of the final post-processing results.In the experiment,the average depth difference of the outer beams deceases from 3.382m to 0.624m and 0.695m after using these two methods.2)Correcton of angular response effect on MBES imagesThe angular response(AR)effect which reduces multibeam backscatter image quality significantly cannot be removed by the traditional AR correction methods effectively.Therefore,the paper puts forward two self-adaptive AR correction methods,namely the AR modeling method and the AR clustering method.The former uses the AR parameters of the average AR curve for modeling and correction,while the latter uses the AR curve of each seabed sediment for AR correction.Both methods overcome the defects of the traditional methods,and improve the image quality obviously.The validities of the methods have been verified by the experiments and the quality of the MBES image has been improved.In the experiment,the average echo strength difference in the specular region deceases from 3.98dB to 0.38dB and 0.17dB after using these two methods.3)Correction of radiation distortion in SSS images in considering the sediment variationAffected by time vary gain residual,beam pattern,angular response and sonar altitude variation,radiometric distortion degrades the image quality of side-scan sonar(SSS)and affects applications of the image seriously.However,the existing methods cannot correct the distortion well,especially when with seabed sediment variation.Therefore,this paper proposes a new radiometric correction method of SSS images with considering seabed sediment variation.Firstly,the characteristics of the different effects on the backscatter strengths(BS)are analyzed and the along-track distortion is removed by building the linear relationship between the distortion and the sonar altitude.Then,based on the fact that the angle-related effects on BSs with the same incident angle are same,a novel unsupervised sediment classification on the SSS image is proposed.Finally,the angle-BS curves of different sediments are gotten,and the angle-related radiometric distortion is corrected.The validities of the proposed method have been verified by the experiments.In the experiment,the average difference of co-located echo strengthes of different survey lines deceases from 17.8dB to 3.6dB.4)Superimposing the SSS images on the MBES bathymetric terrainThe MBES images have accurate position information but low resolution especially in the deep sea,while the SSS images have higher resolution but lack of position accuracy due to its operating mode.Since both of them can reflect the seabed sediment and distribution and have strong complementarity,we can obtain high-resolution accurate-position seabed topography and physiognomy by the integration of the two kinds of sonar images.Therefore,this paper proposes a method of superimposing the side-scan image and the bathymetric terrain based on multibeam and side-scan sediment image matching in each subarea.Firstly,the multibeam image and the side-scan image are processed by angular normalization,k-means++ classification and morphological processing,and the sediment classification image of the two images is obtained.Then,based on the matching points of the sediments and their features in the two kinds of sediment classification images,the position of the side-scan image could be corrected using the rigid transformation in each subarea.Finally,by superimposing the side-scan image with the accurate position on the bathymetric terrain,the high-resolution accurate-position seabed topography and physiognomy can be obtained.Experimental results show that the method is effective and correct.In the experiment,the maximal position offset of the feature points in the SSS image deceases from 19m to 3m.