No abstract

This paper details a methodology for using structured light laser imaging to create high resolution bathymetric maps of the sea floor. The system includes a pair of stereo cameras and an inclined 532nm sheet laser mounted to a remotely operated vehicle (ROV). While a structured light system generally requires a single camera, a stereo vision set up is used here for in-situ calibration of the laser system geometry by triangulating points on the laser line. This allows for quick calibration at the survey site and does not require precise jigs or a controlled environment. A batch procedure to extract the laser line from the images to sub-pixel accuracy is also presented. The method is robust to variations in image quality and moderate amounts of water column turbidity. The final maps are constructed using a reformulation of a previous bathymetric Simultaneous Localization and Mapping (SLAM) algorithm called incremental Smoothing and Mapping (iSAM). The iSAM framework is adapted from previous applications to perform sub-mapping, where segments of previously visited terrain are registered to create relative pose constraints. The resulting maps can be gridded at one centimeter and have significantly higher sample density than similar surveys using high frequency multibeam sonar or stereo vision. Results are presented for sample surveys at a submerged archaeological site and sea floor rock outcrop.

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.

No abstract

No abstract