RGB-D Visual Simultaneous Localization and Mapping (SLAM) Application

Abstract

Simultaneous localization and mapping (SLAM), as one of the important techniques for navigation, robotic mapping and odometry, has been paid a lot of attention by both academia and industry. Visual slam, as an important branch of slam, which relies on visual sensors, such as monocular cameras, stereo rigs, RGB-D cameras, for state estimation, may become the next trend of SLAM due to its price advantages than traditional sensors. In this thesis, we propose a visual SLAM algorithm based on RGB-D camera and deploy it on a real race car (F1/10). The experiment results show our algorithm works well for both online dataset and real environment.