MOBILE ROBOT NAVIGATION USING MONOCULAR VISUAL−INERTIAL FUSION

Jianxian Cai, Penggang Gao, Yanxiong Wu, and Zhitao Gao

Keywords

Monocular visual navigation; sensor fusion; B-spline curve

Abstract

Mobile robot visual navigation technology plays an important role in human production and life. However, it faces many urgent problems. There is a scale problem for monocular simultaneous location and mapping (SLAM) where the map built by monocular SLAM cannot be used by the robot for navigation. In this work, we propose a new method to solve the scale factor using a monocular visual–inertial loose coupling. It can be transformed into an optimization problem and we solve it by nonlinear optimization. Through experiments, we can find that the scale factor is solved by our method. Besides, in the global path planning module, the B-spline curve is used to smooth paths. During it, two control points are added to make the path pass the current location to the destination location. Experiments show that our method is effective.

Important Links:

Go Back