基于约束运动水下机器人视觉悬停研究

Visual station keeping based on constrained motion for underwater robot

悬停能力是水下机器人应具备的重要功能之一,本文介绍基于约束运动的水下机器人视觉悬停方法,该方法利用平面单应矩阵估计潜器的运动参数,并利用潜器的运动约束消除运动估计的多解,仿真和真实水下悬停实验验证了该方法具有较好的鲁棒性和跟踪性能。

Station keeping is one of the important function which underwater vehicle is required. This paper introduces a vision-based station keeping method for underwater vehicle, which exploits plannar homography for underwater vehicle motion parameters estimation. Multiple solutions are eliminated by constrained motion of the underwater vehicle. This method is validated by simulation and real image experiment.