Costmap Construction and Pseudo-Lidar Conversion Method of Mobile Robot Based on Monocular Camera

摘要

With the development of robotics and computer vision technology, monocular cameras with low cost advantage are widely used in mobile robots. In the research of mobile robot obstacle avoidance competition based on ROS and monocular RGB camera, this paper does not directly use monocular camera for navigation and obstacle avoidance, but proposes a novel costmap creation and pseudo-Lidar transformation method based on monocular vision. The main work of this paper is to process monocular vision images, convert the image coordinate system to map coordinate system through camera calibration, display obstacle information on the costmap, extract obstacle edges and publish them in Lidar data format. This paper realizes the conversion of monocular vision to map and Lidar, completes the environment perception and mapping of mobile robots, and makes it possible to use Lidar-based navigation algorithms and toolkits.