Autonomous Driving is one of the megatrends of today. However, many challenges are still unsolved. One of these challenges is the localization and motion estimation of vehicles. Typically, this is achieved with some prior available environmental data. Unfortunately, this kind of information is not always available. Still, the vehicle must retain a way to cope with unknown environmental situations. That is where self-contained ego- and object motion estimation techniques come into play. These methods can be based on various sensor types. The goal of this thesis is to evaluate, implement and compare different approaches for the use case of a self-driving model car.
«
Autonomous Driving is one of the megatrends of today. However, many challenges are still unsolved. One of these challenges is the localization and motion estimation of vehicles. Typically, this is achieved with some prior available environmental data. Unfortunately, this kind of information is not always available. Still, the vehicle must retain a way to cope with unknown environmental situations. That is where self-contained ego- and object motion estimation techniques come into play. These met...
»