Optical inspection of ports and harbors: laser-line sensor model applications in 2 and 3 dimensions

Abstract

There are 361 ports of interest to the US Coast Guard regarding homeland security issues. Speed and accuracy of inspections there for "foreign objects" is critical to maintaining the flow of commerce through these ports. A fusion of acoustic and optical imaging technologies has been implemented to rapidly locate anomalies acoustically and inspect them optically. Results of field tests are presented. Effective deployment of AUV-or ROV-mounted optical sensors to inspect ship hulls and port facilities will depend on accurate, real-time prediction of the sub-surface optical environment and upon accurate sensor models parameterized for the time and place of inspection. For bi-static laser-line scanner sensors such as the Real-time Ocean Bottom Optical Topographer (ROBOT), ambient light decreases the range to the inspection object (e.g. hull) for which laser-line contrast is adequate for ranging and imaging in 3-D. Reduced range implies narrower swaths and longer inspection times. A 2-D and 3-D hybrid marine optical model (HyMOM) of the environment beneath ships or adjacent to sea walls and pilings has been developed, applied and validated in eutrophic and mesotrophic settings, and a Monte Carlo sensor model of ROBOT has been developed. Both are discussed and combined to evaluate sensor performance in different environments. To provide the inherent optical properties needed to run such models, data from the Autonomous Marine Optical System (AMOS) were collected and transmitted back to the laboratory. Examples of AMOS results and model outputs are presented.

A Focus on Recent Developments and Trends in Underwater Imaging

“…This configuration significantly reduces backscatter in the raw data and enables recovery of the scene's 3-D structure by means of triangulation (Dalgleish et al, 2005;Carder et al, 2005). Narasimhan et al (2005) present two innovative additions to this method.…”

“…Figure 10 shows one such ROV-mounted system that combines DIDSON, a video camera and parallel lasers for fish stock assessments and quantification (Yamanaka et al, 2008;Yamanaka, 2005). Other applications well suited to using DIDSON (and optionally a conventional camera system) include close-range inspection of manmade structures such as dock and bridge pilings (Kloske, 2005) (see Figure 11), ship hulls (Negahdaripour and Firoozfam, 2006;Kloske, 2005) (see Figure 12) and oil pipelines to name a few.…”

Model-based Hyperspectral Exploitation Algorithm Development

“…In cases where there were significant discrepancies, it is believed that the implemented model represented a more sophisticated treatment of the problem. Comparisons were also demonstrated against Carder et al (2005) and Zaneveld et al (2001) where spatial obscuration and surface effects were simulated showing agreement to the modification to the light field under these conditions. In order to achieve the correct radiometry of these simulations, substantial modifications have been made to the DIRSIG simulation framework to develop a more general and flexible representation of radiometric concepts/elements and their interaction across the different media encountered in the littoral environment.…”

Underwater 2D Image Acquisition Using Sequential Striping Illumination

“…As one of the essential 3D measurement technologies, structured lighting is increasingly being applied in underwater imaging; those frequently required include, seabed mapping, pipeline or dock detection, target localization, and ROV (remotely operated vehicles)/AUV (autonomous underwater vehicles) navigation. A variety of structured lighting techniques have been developed for these applications, such as laser line scan systems [1], 3D laser line scan mapping systems [2], laser line sensors [3], and seabed-relative navigation by structured lighting techniques [4]. In general, structured lighting techniques modulate the appearance of a surface by projecting a particular pattern of light onto it [5]; subsequent correspondence of the image pattern and surface points, provided with calibration of the image sensor and projector, enables accurate computation of the range of surface points.…”