A new method for artefact-free estimation of surface slope from bathymetric Lidar data

When estimating the slope and aspect of a natural surface obtained by bathymetric Lidar scanning, discrepancies in the elevation data from neighbouring strips often cause artefacts. In this paper, a novel algorithm for slope estimation avoiding this kind of artefacts is presented. The algorithm is based on filtering the point data in the region of overlap using a set of scan angle based thresholds. Each threshold yields a data set with different selection of points from the neighbouring strips. Gradient estimates based on these data sets are then combined by either averaging or applying a trimmed mean type operation to obtain the artefact-free slope estimate. The algorithm is developed using bathymetric Lidar data and the obtained slope estimate of the seabed is used for the correction of the Lidar waveform data. The developed method is applicable in a wide range of situations where overlapping data from different sources need to be combined.

在对测深激光雷达（bathymetric Lidar）扫描获取的自然地表开展坡度与坡向估算时，相邻航带的高程数据差异常会引发伪影（artefacts）。本文提出了一种可规避此类伪影的新型坡度估计算法。该算法通过一组基于扫描角的阈值，对重叠区域内的点数据进行滤波筛选。每一组阈值都会从相邻航带中选取不同的点集，进而生成对应数据集。随后，基于这些数据集得到的梯度估算结果，可通过均值平均或截尾均值类运算进行融合，最终得到无伪影的坡度估算值。本算法基于测深激光雷达数据开发，所得海底坡度估算结果可用于激光雷达波形数据的校正。所提出的方法可广泛应用于需融合多源重叠数据的各类场景。