Passive Optical Sensing for Automatic Stationkeeping of ROVs

在过去的十年中，人们对水下机器人车辆在各种海洋应用中的需求不断增加，例如管道铺设，检查，维修，地质调查等。可靠性、安全性、灵活性和经济性是智能远程操作车辆（rov）发展的主要因素之一。这允许更有效地使用车辆、操作员和他们的支援船。最终目标是实现具有在各种水下任务中自主运行所需的所有智能的系统。在ROV系统中建立一定程度的自主性的一个基本能力是准确地感知相对于固定坐标系或附近物体的位置和运动。随着这些车辆得到更广泛的使用，一种特殊的应用是近底部的驻地维持；也就是说，在海底附近保持一个理想的位置来执行一些任务。在目前的技术下，近海底定位可以通过主动传感来实现，使用声波或光学信标，这些信标需要精确地定位在海底，并在任务完成后恢复。在避免主动传感的应用中，这是昂贵的，繁琐的，并且非常不切实际的。这个为期三年的项目涉及到对这个问题的另一种解决方案的研究，即通过利用底部自然特征作为信标来进行被动光学传感。目标是开发计算机视觉技术，通过解释海底环境的图像来提取有关车辆位置和运动的信息。

Last decade has seen an increasing demand for use of underwater robotic vehicles for various ocean applications, for example pipeline laying, inspection, repair, geological survey, etc. Reliability safety, flexibility, and economics are amongst some of the major factors that demand the development of intelligent Remotely Operated Vehicles (ROVs). This permits more efficient use of the vehicle, operator, and their support ships. The ultimate goal is the realization of systems with all the intelligence required to function autonomously in a variety of undersea tasks. One general capability essential to building some degree of autonomy into the ROV systems is accurate sensing of position and movements relative to a fix coordinate system or nearby objects. A particular application, as these vehicles become more widely used, is near-bottom stationkeeping; that is, maintaining a desired position to perform some task near ocean floors. With the current technology, near-bottom stationkeeping can be done via active sensing, using acoustic or optical beacons that need to be positioned accurately on the ocean floor, and recovered upon completion of the task. This is costly, tedious, and highly impractical in applications where active sensing is to be avoided. This three-year project involves the investigation of an alternative solution to the problem, passive optical sensing by exploiting the bottom natural features as beacons. The goal is to develop computer vision techniques for extracting information about position and motion of the vehicle by interpreting images of the undersea environment.