An Improved Multi-Beam Trend Surface Imaging Method Based on Surface Curvature Compensation

Abstract

For multi-beam trend surface imaging, a polynomial is typically created according to the least squares method to fit a surface similar to the actual data. However, in the filtering process, the fitting result will be different from the real situation due to the particularity of some seabed planes. In order to make the surface fit closer to the actual terrain, the curvature correction can be introduced to correct the error. On the basis of comparing the local curvature of the fitted surface with the local curvature of the actual data, a new curvature correction is obtained, and a new trend plane filtering method is proposed by using the new curvature correction. The effectiveness of the new method is verified by MATLAB simulation. The effect is shown in Figure 2. This method can better preserve the local characteristics of the surface.