Description

This technology represents a device and method for measuring seafloor (or other objects where topographical representation is desired) roughness, using low-power lasers and a video recorder, that may be deployed on a remotely operated vehicle (ROV) and other platforms with similar limitations in size, weight, and power. Development began as an initiative to understand how sound waves behave for different sea floor characteristics which cause scattering of the sound waves. The technology utilizes six evenly spaced lasers with conical lenses to disperse the beams in a planar fashion along the measured area. A video camera is directed at the measured area on a 45-degree angle from the laser mounting surface. This enables the video to capture all six lasers as the system traverses over different surface profiles. The captured data is then processed using simple geometry, where each pixel represents a straight line to the camera and each laser represents a plane. The intersections are calculated and a surface profile is generated

Back-end processing enables the system to map the surface topography to the captured images as well as determine the standard deviation for the group of six lasers. Having six lasers provides redundant measurements for positional data, enabling the system to self-check itself and validate any areas of concern

The purpose of data processing is to produce a measurement of the fine-scale seafloor bathymetry. The process may be divided into a number of steps

1. Extraction of 2D image pixel coordinates of the laser stripes from each video image

2. Projection of the 2D laser stripes from step 1 into 3D space referenced to the camcorder, using pitch and roll data and the known relative positions of lasers and camcorder: This is a simple geometry problem in which each pixel of the video image represents a ray emanating from the optical center of the imaging sensor

3. Computation of camcorder trajectory and fine-scale heading over the seafloor, using the altitude information from step 2 and vehicle heading sensor data, by cross-correlation of seafloor images projected onto a common seafloor plane. To reduce it to a simple 2D correlation process, the scale and orientation of the images must be constant. This was achieved with the pitch and roll sensor data, the height above bottom estimate from the 3D laser stripes, and the heading sensor data

4. Construction of the seafloor surface by stacking the 3D stripes from step 2 using the trajectory and heading data from step 3, and iteratively reducing the residual error by adjusting fine-scale pitch and roll angles to achieve a least-squares-error solution. This process requires accurate depth information which was not available.

Benefits

• Low cost alternative to current technologies
  
• Increased accuracy at 1mm x 2mm spatial resolution
  
• Deployable from an ROV
  
• Capable of self-checking accuracy of measurements
  

Market Potential/Applications

• Seafloor topography mappin
  
• Ship hull inspection
  
• Underwater pipe and cable inspections
  

Contact:

University of Texas,
Austin, USA
Website : www.otc.utexas.edu