Computer vision system for an autonomous underwater vehicle

摘要

This thesis describes a vision system for the PISCIS project. It is a proposed project for the use of an untethered autonomous underwater vehicle (AUV) for pipeline inspection. The vision system is termed PVS (the PISCIS vision system). It will assist the AUV in finding and following pipelines. The PVS will use a camera to continuously find the position and heading of pipelines relative to the AUV. A salient feature of the PVS is that it is designed to find all pipelines within the field of view and, thus, the AUV can follow any of them. The positions and headings are used in a control loop to move the vehicle to a chosen pipeline. They are also used to keep the vehicle over this pipeline during the inspection. Robust image interpretation is important as humans can not interact with the PVS and correct errors. Furthermore, the image quality is reduced by backscatter from the water, light absorption, a non-uniform background (the sea-bed), and marine material on the pipelines. The PVS is fixed to a moving vehicle and must rely on a heading sensor and an altitude sonar to match images features with pipeline models. The models are retrieved from a map based on the vehicle's position. The PVS is tested with images taken in air in the laboratory. The images have, like underwater images, highly varying contrast, ''pipelines'' with a broken outline, and a non-uniform scene background. The conclusion is that the PVS can interpret moderately complex pipeline-configurations even when the vehicle's position is inaccurate. A real-life system must function over a wide range of image acquisition conditions; the PVS is shown to be robust against vehicle pitch-, roll-, and heading-errors, scaling errors, and a skewed light source. 99 refs., 83 figs., 14 tabs.