Highly detailed mapping and precise localization have boosted the performance of self-driving cars and have become an important component of L3+ vehicle automation. HD maps not only augment the information acquired by sensors, but lift environment perception to a telematic horizon, which extends far beyond the sensors’ field of view. This module will provide an overview on fundamental concepts and challenges of HD mapping and localization techniques in theory and practice. It will elaborate offline and online approaches to map building and localization-only methodologies. Practical issues like map validation and crowd mapping concepts will be addressed. Throughout the module, practical applications for automated driving will be discussed.
What You’ll Learn
- Gain an overview on the content, formats and accuracy requirements of HD maps and localization therein
- Understand the basic theory of simultaneous localization and mapping (SLAM) techniques for automated mapping
- Formulate measurement constraints and apply SLAM fusing different sensor types yourself for SLAM and localization-only applications
- Understand concepts for map validation and crowd mapping
- Know practical applications for map-based automated driving
Module Content
Module Kick-Off
Introduction
Fundamental Theory
Simultaneous Localization and Mapping
Applications and Extensions
