This paper introduces a vehicle safety system for moving backwards during night time. The system consists of two RGB-D cameras supported by two computers connected via Gigabit Ethernet. The cameras are mounted on the back side of the vehicle at different altitudes and tilt angles allowing detection of objects behind the vehicle and defects of the road surface. The safety system triggers an alarm if an object appears in dangerous proximity to the car. The dangerous proximity is decided according to the mass of the vehicle and its speed. Obstacles are tracked to predict if ones may become dangerous in the near future. Experiments have shown that the system is able to detect obstacles and holes on the road pavement when there is not enough light for common cameras.
CITATION STYLE
Filonenko, A., Hernández, D. C., Vavilin, A., Kim, T., & Jo, K. H. (2015). Distance sensor fusion for obstacle detection at night based on kinect sensors. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9226, pp. 135–144). Springer Verlag. https://doi.org/10.1007/978-3-319-22186-1_13
Mendeley helps you to discover research relevant for your work.