Visual-Pressure Fusion for Underwater Robot Localization With Online Initialization

Abstract

The motion of underwater robot is usually slow, which leads to large scale error easily introduced in the positioning method based on monocular Visual Inertial Odometry (VIO). To solve the above problem, we present a positioning method based on visual-pressure fusion in this letter. Firstly, it is proved by observability analysis that when the robot's motion is not on the same horizontal plane, the system scale can be uniquely determined by monocular image and depth information; secondly, a novel visual-pressure initialization method is proposed, which can calculate the system scale and initial attitude with respect to world frame online; finally, we integrate our method into current advanced monocular Simultaneous Localization And Mapping (SLAM) system, and the effectiveness and superiority of the method are verified by experiments on simulation datasets and real datasets.