Towards Accurate 3D Road Models from Radar Detections for Advanced Driver Assistance Systems

Félév: 2023-2024 tavasz

Kategória: Nincs megadva

Téma leírása

Advanced Driver Assistance Systems (ADAS) play a pivotal role in enhancing vehicle safety and enabling autonomous driving. One of the fundamental requirements for ADAS is the accurate perception of the road environment, especially in scenarios with poor visibility or challenging weather conditions. This topic aims to investigate and develop a methodology for creating accurate 3D road models using radar detections. By fusing data from radar sensors, it becomes possible to enhance the perception capabilities of vehicles, ultimately contributing to safer and more reliable autonomous driving.

Objectives:

• Radar Data Preprocessing:
• Investigate methods for preprocessing radar data, including noise reduction, clustering, to extract meaningful information about the road geometry.
• 3D Road Reconstruction:
• Design algorithms to convert processed radar data into a detailed 3D representation of the road surface, including road boundaries, road curvature, lane topologies and elevation changes.
• Validation and Accuracy Assessment:
• Devise methodologies to validate the accuracy of the generated 3D road models against ground truth data

 

To solve the task, the student receives help from the staff of the Continental AI Development Center.

Külső partner: Continental Autonomous Mobility Hungary

Maximális létszám: 1 fő