Verification of the necessity of updating the Vector-Map used for positioning estimation
Yujiro KANAMORI, Ryuichi YOSIE, Junichi MEGURO · The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec) · 2024
In recent years, research on autonomous driving has been conducted actively. In particular, with concerning self-position estimation, technologies that use 3D point cloud maps created from LiDAR scan data and Vector-Map, which stores information such as road surface conditions and signs as vector data, are attracting attention. Therefore, in this study, we examine the necessity of updating Vector-Map when there are environmental changes such as the redrawing of white lines. Therefore, we investigate the environmental changes that occurred and create a Vector-Map before and after the environmental changes based on those changes. By using sensor simulation and acquiring sensor data, we will evaluate the accuracy of position estimation before and after environmental changes and consider the need for map updates.