About
CycleSafely is an innovative bicycle safety system that uses computer vision, LIDAR sensors, and machine learning to detect and track vehicles around cyclists. The system measures distances to moving vehicles, analyzes their trajectories, and provides real-time alerts about dangerous situations such as close passes and potential collisions. Project leader is Stefan Ohrhallinger
Cycling in the city poses significant risks from close passes and dangerous interactions with motor vehicles. CycleSafely addresses this challenge by combining RGB-D cameras, LIDAR sensors, GPS, and IMU data to create a comprehensive safety monitoring system that can detect hazardous situations and alert both cyclists and drivers in real-time.
Key Technologies: YOLOv4 for vehicle detection, DeepSORT for tracking, ORB-SLAM3 for registration, GPS+IMU sensor fusion, Livox Mid-360 LIDAR, and neural network-based trajectory prediction.
Get Involved as a Student
Are you a student interested in contributing to bicycle safety research? We welcome participation from motivated students through project work, bachelor's theses, or master's theses. Learn more about opportunities to join the CycleSafely project: