Implementation and research on indoor mobile robot mapping and navigation based on RTAB-Map

The positioning, mapping and navigation of mobile robots has always been a hot topic which attracts widespread attention around the world. In this paper, mapping and navigating in a complex indoor environment of 100 square meters is done with the RTAB-Map algorithm for multi-sensor information fusion using an RGB-D camera, a single-line lidar, a gyroscope, an odometers and other sensors. Navigation tasks will be published after the map construction tasks is done, the experimental results are scored based on expressions of different approach. This study compares the scene data collected with the data in the real environment, analyzes the accuracy and error of the map with different methods. This study also compares the visual-laser combined mapping and navigation scheme with the pure vision schemes or pure laser schemes. Result shows that the task completion is 100% with the visual and lidar combined, compared with the scheme that uses lidar only (45%) or the RGB-D camera only (75%).