The industry's recognized authority for design, engineering and application of equipment and services in the global ocean community
Issue link: http://sea-technology.epubxp.com/i/86750
3D Seafloor Mapping With Automated Data Analysis The Generation and Application of 3D Color Reconstructions For Quantitative Algorithm-Based Analysis By Adrian Bodenmann Project Researcher Blair Thornton Associate Professor and Tamaki Ura Professor Underwater Technology Research Center Institute of Industrial Science The University of Tokyo Tokyo, Japan I maging of the seafloor has been one of the ma- jor applications of underwater robots over the past decade. Different methods of combining individual photos into continuous maps have evolved, and methods to process images in 3D are becoming more common, making it pos- sible to visualize lifelike reconstructions of the bottom of the ocean. While 3D reconstructions provide a natu- ral medium to show the seafloor and present a reference to overlay data measured by other sensors, these data can also be directly analyzed by algorithms to extract scientifically useful information. A method based on laser scanning, which generates high-resolution bathymetry maps, was chosen and extended to match color information to generate 3D reconstructions. This approach is relatively simple in terms of its hardware and software requirements because the required images can be obtained by a single camera, and all calculations are feed-forward with no feature-based matching. Hardware and Mapping Method The underwater vehicle, which can be an AUV or ROV, is equipped with a camera, a sheet laser, lights and navigation sensors such as a Doppler velocity log (DVL) with an inertial navigation system (INS) compass and a depth meter. A com- puter in a pressure-tight housing saves time-stamped images from the camera and logs the navigational measurements. The AUV Tuna-Sand with the equipment for recording data to generate 3D reconstructions. The lights are directed to illuminate only a part of the camera's field of view, and the sheet laser projects a laser line in the unilluminated area, perpendicular to the forward direction of the vehicle. The camera is mounted at an angle and with an offset to the sheet laser, and the images taken of the laser line reveal the topography of the seafloor. The algo- rithm analyzes all images in post-processing and extracts the laser line with subpixel resolution, based on which shape of the seafloor is calculated. Each image provides one line of bathymetry points, and by combining lines of data from multiple images taken while the vehicle moved forward, a bathymetry map of the scanned area is generated. In the second step, the color of each bathymetry point is determined. To do so, a corresponding image is selected where the spot on the seafloor appears in the illuminated section of the image based on the vehicle's navigation data. www.sea-technology.com OCTOBER 2012 / st 41