FFT-based Terrain Segmentation for Underwater Mapping

Abstract

Abstract-A method for segmenting three-dimensional scans of underwater unstructured terrains is presented. Individual terrain scans are represented as an elevation map and analysed using fast Fourier transform (FFT). The segmentation of the ground surface is performed in the frequency domain. The lower frequency components represent the slower varying undulations of the underlying ground whose segmentation is similar to de-noising / low pass filtering. The cut-off frequency, below which ground frequency components are selected, is automatically determined using peak detection. The user can specify a maximum admissible size of objects (relative to the extent of the scan) to drive the automatic detection of the cut-off frequency. The points above the estimated ground surface are clustered via standard proximity clustering to form object segments. The approach is evaluated using ground truth hand labelled data. It is also evaluated for registration error when the segments are fed as features to an alignment algorithm. In both sets of experiments, the approach is compared to three other segmentation techniques. The results show that the approach is applicable to a range of different terrains and is able to generate features useful for navigation.

Experiments on Surface Reconstruction for Partially Submerged Marine Structures

“…Another way to capture the numerical quality of the accuracy of the registration is described by Douillard et al. (). To show consistent reconstruction from both parts of the marine structure of interest, we present an occupancy grid‐based map of the above‐ and below‐water parts of the structure and the probability of each cell to be occupied. The color indicates the probability of the cells to be occupied, Figure .…”

“…Another way to capture the numerical quality of the accuracy of the registration is described by Douillard et al. ().…”

Benthic monitoring with robotic platforms — The experience of Australia

“…Underwater environments are typically unstructured and present substantial challenges to traditional approaches for segmentation and classification of objects. We have started addressing some of these problems in the context of separating the ground from objects in underwater scenes [42] (Figure 9). 3) Adaptive sampling: Bathymetric maps derived from sonar observations provide relatively low-resolution maps of the seafloor.…”

Autonomous Landing of Underwater Vehicles Using High-Resolution Bathymetry

“…However, these previous works do not develop the sensing and data processing methods needed to automatically identify areas where a vehicle can land safely. Regarding methods to analyse seafloor terrains, Fourier analysis based segmentation and 3D alignment was described in [26], [27]. Other methods for surface classification using wavelets [28] have also been described, but have not been applied to landing site identification.…”