Mapping performance comparison of 2D SLAM algorithms based on different sensor combinations

Abstract

Currently, some existing SLAM algorithms were widely applied in the field of mobile robots. However, most SLAM algorithms were tested either by simulation or using offline bags storing messages, which ignored other contributing factors influencing mapping including the changes of position and speed of the robot during it autonomous movement. The aim of this study is to investigate the solution of constructing more accurate map with high quality sensors based on freely available SLAM algorithms. Comparisons and analyses of three 2D SLAM algorithms (i.e. Gmapping, Hector SLAM and Cartographer) available in ROS were conducted to map different environments in our work using different sensor combinations of a LIDAR, a Stereo camera and an IMU, respectively. The research showed that the Cartographer algorithm has advantage over two other algorithms on constructing well-defined and informative environmental maps with minimal errors. Also, the LIDAR scan is more suitable than the ZED scan as the input of mapping algorithms, and the Cartographer combining the LIDAR scan, ZED odometry and IMU is the best solution to map in all combinations.