28th Aachen Colloquium Automobile and Engine Technology 2019
AI Based Method of Precise Trajectory Extraction from Imprecise Open Road Data
Authors
Jan Olšina, Miroslav Zima, Cesar Lobo-Castillo, Pavel Jiroutek, Valeo Autoklimatizace k.s., Praha
Summary
The scope of this paper is to describe the precise process of trajectory extraction from vehicle odometry, lidar scanpoints and GNSS. It combines a state of the art machine learning method for automated extraction of road boundary with an advanced method for fitting lidar measurements to road boundary data and extraction of the compensated trajectory. Experimental results and performance on a large dataset are presented, as well as a comparison with other contemporary methods. One of the main benefits of the outlined approach is its robustness against errors resulting from integration and imprecise odometry measurement. This approach also allows for direct output of accurate pre-annotated road boundaries in a global coordinate frame which can be directly used in later system performance evaluation.
Members of the Austrian Society of Automotive Engineers have access to all lectures of the International Vienna Motor Symposia.