Experimental Research on Feature Extraction of Laser SLAM Based on Artificial Landmarks

摘要

Simultaneous Localization and Mapping (SLAM) is the core issue in the field of mobile robots. Laser SLAM is one of the widely used solutions for the application of SLAM in engineering. Laser SLAM estimates the change of the robot pose using the features of the environment. However, it is difficult to extract natural features in a highly similar environment. Therefore, artificial landmarks are needed to be placed in the environment when necessary. In this paper, the reflective characteristics of the artificial landmark are analyzed experimentally, and two feature parameters for the extraction of the artificial landmark are proposed based on their reflective characteristics. One of them is the intensity of the reflected light received by the laser ranger, and the other is the number of the laser points that the artificial landmark can continuously reflect back in each scan. Through the statistical analysis of the laser data collected in the experiment, the specific values of these two parameters are determined, and the detection range of the laser ranger to effectively extract the landmark under the selected parameters is measured experimentally. These results provide experimental support for the deployment of the landmarks in the environment.