| Multibeam bathymetric system is an important tool for sea exploration recently. There are all sorts of interference, for the reasons of both human and nature, in the measurement of multibeam data inevitably. The depth of these disturbances will bring the quality of submarine topography, must be removed. In multibeam bathymetry data processing, it is the main task to choose suitable methods, in order to eliminate errors,get an exact measuring result, and enhance the quality of the digital chart.In this paper some key technology of multibeam bathymetric data processing are studied:Firstly, this paper analyzes the steps of the system of multibeam information's processing and the format of original data stored by the multibeam bathymetry system and supplies a method on how to extract the original data.Secondly, the author proposes a new method called bilateral exponential weighting method to improve the normal exponential weighting method, and uses Kalman Filtering Algorithm, aiming to eliminate gross erros in vessel attitude data. Besides, two methords are used for track fitting.Thirdly, this paper analizes some common criterions for detecting gross errors, introduces two algorithms-median fitering and distance weighted median filtering. From the characteristic of multibeam bathymetric data, a new median fiter based on universal gravitational theory is proposed, which is better than the other median fiter reffered.Fourthly, a modified model-cubic spline interpolation is presented based on relevant maths theory.Through the comparison, this new method overcomes the shortages of the trend-surface filter and is effective.Fifthly, a new approach of editing the depth data based on Support Vector Machine(SVM) is discussed, and the implantation of construction of denoising filter by applying three kinds of kernel functions are discussed as well.Thourgh comprehensive consideration during these kernel functions, Gauss kernel function is better than the others . |
PDF Full Text Request