Fusion of Radar- and Lidar-Data for Object-Tracking-Applications at Feature Level

The topic of automation can be found in almost every area these days. In Industry 4.0 in particular, the exclusion of hazards to people, for example in a human-robot collaboration, is a central issue. Object tracking is an option for workspace monitoring in this context. To improve object tracking, a fusion is to be designed and implemented in this work in order to combine the various measurement properties of a Radar- and a Lidar-sensor. In this paper, the Kalman filter and the fusion in general are briefly introduced. Object tracking, the concept of the implemented fusion and the arrangement and synchronization of the sensors are then described in more detail. A final evaluation leads to the conclusion that especially the availability and the reliability in object tracking can be improved. Furthermore, it can be shown that feature-level fusion has a significant advantage over symbol-level fusion.