Developments in laser-line scanned undersea surface mapping and image analysis systems for scientific applications

摘要

The ability to produce range maps of undersea objects has direct application in a variety of applications including robotic manipulation and control, collision avoidance and navigation, military detection, identification, and localization, and scientific research. Although systems have been proposed using acoustic sensing methodology, optical systems can provide higher resolution and improved discrimination at close range. This paper describes recent developments using a laser scanner to construct accurate range maps of marine snow aggregates at high resolution (200/spl times/200/spl times/200) and at range distances of less than 10 cm. The proposed method is also suitable (with modifications) for long range operation at up to 6 attenuation lengths. The system utilizes a triangulation sensing approach that is suited to low cost implementation through the use of a novel, patented method for obtaining position information. The approach is also compatible, with range gating methods for reduction of scattered light interference, at multiple attenuation lengths. A description of the system theoretical limitations, design tradeoffs, and practical design constraints is presented. The design approach and recent hardware and software implementation developed under NSF and University of Southern Mississippi sponsorship are described. Graphical displays showing underwater surface map images of marine snow are also presented. Examples of information obtainable from the surface map data reduction of marine snow images include range histogram, particle quantification, surface area and volume, and direction and speed of motion. Future goals and modifications for the system are discussed.