The perception of the environment of a driving vehicle by various sensors is one of the most challenging tasks to solve for safety and comfort driving systems. The usage of map data is an important component for a reliable scenario. This paper deals with the processing of multi-layer laser detections to accumulate local 3D environment information for representing them in a suitable map. A multi-level segmentation is applied to the raw measurements and the resulting higher level data. Kalman Filters at several levels are used for both stabilizing the extracted objects and determining whether an object exists or not.
Recursive building of local maps using multi level feature fusion
2007
6 Seiten, 8 Bilder, 5 Quellen
Aufsatz (Konferenz)
Datenträger
Englisch
British Library Conference Proceedings | 1999
|Vehicle Detection Using Multi-Level Probability Fusion Maps Generated by a Multi-Camera System
British Library Conference Proceedings | 2008
|Feature-level fusion in personal identification
IEEE | 2005
|