Benchmark system for novel 3D SLAM algorithms

Abstract

In this days the autonomous navigation and the mapping are getting more and more attention. During the article, a test environment was represented to help the test of a three-dimensional Simultaneous Localization and Mapping (SLAM) method. The program is able to operate more RGB-D sensor based databases. In the program, an own three-dimensional SLAM algorithm was implemented. Test environment specifies the estimated movement of the robot and three-dimensional point cloud of the traveled area. Finally, the accuracy of the method is tested knowing the estimated route and real route belonging to databases. At the end the system is able to export the completed point cloud to make post-processing on it later.