Comparison of Scene Reconstruction Methods Based on Hector SLAM

Rui Pan,Qian Sun,Lei Xing,Fei Peng,Xubo Wang

DOI: https://doi.org/10.1109/ciss57580.2022.9971285

2022-01-01

Abstract:Scene reconstruction, which could help the driver obtain the surrounding information of the vehicle, provides essential information for the in-vehicle system. Furthermore, sensors are fundamental components to collect data for scene reconstruction. In order to make millimeter wave radar more widely used in the automobile industry, this paper compared the scene reconstruction performance of millimeter wave radar with lidar based on Hector SLAM. First, the coordinate system of millimeter wave radar was converted to world coordinate system, which was the coordinate system used in Hector SLAM. Then the data format conversion of millimeter wave radar was completed to meet the requirements of Hector SLAM. Finally, the experiment was designed to make scene reconstruction for cars outside. By using cosine similarity and grayscale histogram to analyze the scene reconstruction performance between millimeter wave radar and lidar, experiment results show that there is little difference between millimeter wave radar and lidar in scene reconstruction based on Hector SLAM. So that millimeter wave radar could have extensive applications in the automobile industry.