Tool for Creating High Definition Maps in Lanelet2 format
Name
Karl-Johan Pilve
Abstract
At the University of Tartu, the Autonomous Driving Lab (ADL) conducts research and development in the field of self-driving vehicles. As modern self-driving vehicles need detailed high definition vector map for driving, the ADL has developed a tool for drawing such maps and mapped the first test tracks on the streets of Tartu and Tallinn. A high definition vector map in Autoware format, which is compatible with software for self-driving vehicles, is generated from ADL's vector map. ADL wishes to start testing with maps in Lanelet2 format because it has many advantages over Autoware vector map format.
This bachelor's thesis aimed to create a tool that would convert the existing ADL's vector map to Lanelet2 format. This thesis gives an overview of high definition maps and Lanelet2 map format. Then the technical implementation is explained, and an overview of problems during the development process is provided. Finally, the description of the testing process and suggestions on improving the tool are also given.
This bachelor's thesis aimed to create a tool that would convert the existing ADL's vector map to Lanelet2 format. This thesis gives an overview of high definition maps and Lanelet2 map format. Then the technical implementation is explained, and an overview of problems during the development process is provided. Finally, the description of the testing process and suggestions on improving the tool are also given.
Graduation Thesis language
Estonian
Graduation Thesis type
Bachelor - Computer Science
Supervisor(s)
Tambet Matiisen, Edgar Sepp
Defence year
2021