The recognition of pedestrians in precrash scenarios or in other road traffic situations is a very challenging task. Depending on the applied sensor system, the recognition consists of a detection of obstacles and a classification algorithm in order to separate pedestrians from non-pedestrian objects. Amongst others, a Photonic Mixer Device (PMD) system exists which is able to recognize many pedestrians in different traffic situations. This paper presents the improvement of an existing PMD based classification system using a Mono Video (MV) grayscale camera and appropriate image processing algorithms additionally. The resulting system is called Pedestrian Recognition System Using a Multi-Sensor Environment (PReSUME). It is able to perform the classification task with an increased robustness while still providing a relatively high update rate.
Improvement of a PDM based classification of pedestrians in road traffic scenarios
Verbesserung einer PDM basierten Einklassifizierung von Fussgängern bei Strassenverkehrsszenarien
2008
6 Seiten, 4 Bilder, 2 Tabellen, 11 Quellen
Aufsatz (Konferenz)
Englisch
DEVICE AND METHOD FOR PREVENTING ROAD PEDESTRIANS FROM CROSSING TRAFFIC
Europäisches Patentamt | 2023
Road traffic incidents involving pedestrians in areas with limited visibility
DOAJ | 2024
|Detection of the human skeleton on pedestrians in road traffic
VDE-Verlag | 2022
|DEVICE AND METHOD FOR PREVENTING ROAD PEDESTRIANS FROM CROSSING TRAFFIC
Europäisches Patentamt | 2022
|Pedestrians and traffic signals
Engineering Index Backfile | 1966
|