This thesis presents methods for vehicle localization within known environments like car parks and underground car parks. These methods enable autonomous parking solely using the cars series production sensors. On the one hand, a particle-filter-based approach was developed, which estimates the vehicle’s pose using ultrasonic sensors, vehicle’s odometry and an a priori map of the environment. On the other hand, a camera-based approach was realized, which performs vehicle localization using image processing to detect semi-natural (here: car park markings) and artificial landmarks (here: AprilTags).
«
This thesis presents methods for vehicle localization within known environments like car parks and underground car parks. These methods enable autonomous parking solely using the cars series production sensors. On the one hand, a particle-filter-based approach was developed, which estimates the vehicle’s pose using ultrasonic sensors, vehicle’s odometry and an a priori map of the environment. On the other hand, a camera-based approach was realized, which performs vehicle localization using image...
»