Advanced camera and LiDAR localization methods for automotive scenarios
With the rapid advancements in autonomous vehicle technologies and the latest Advanced Driver Assistance Systems (ADAS), precise and reliable localization has become critical, particularly in challenging environments where traditional GNSS-based solutions fall short. This research project focuses on developing innovative 3D localization solutions to address these challenges using different sensor modalities, offering complementary approaches to achieve the same goal. ; ; The first method investigates LiDAR map-based localization, aiming to design a highly accurate algorithm focusing on efficiency that integrates with an Extended Kalman Filter, forming a robust localization stack for real-time vehicle positioning in GNSS-denied environments. The second method explores camera-based localization, presenting a more easily implementable solution that excels in compact and feature-scarce settings. This method supports a broader range of applications, including Augmented Reality for in-vehicle systems, and is developed with the goal of enhancing ADAS capabilities for drivers.
Back to Current Students