Multi-beam Data Automatic Filtering Technology

Abstract

Aiming at the characteristics of complex noise sources in multi-beam bathymetric data, this paper proposes a multi-beam automatic filtering method that combines filtering of optimal reference curved surface and trend surface. By implementing statistical filtering to pre-process the data, the optimal reference curved surface is constructed based on the filtered terrain data, the theoretical optimal depth value of each beam point is calculated. The optimal reference curved surface is filtered by combining the depth tolerance to determine whether the point is a noise point. Then the trend surface is constructed using the filtered non-noise data, and the trend surface is filtered on the original point cloud data. Through the verification of the measured data, this method can effectively remove most of the cluster noises in the multi-beam bathymetric data and is more efficient than manual processing.