Title :
Stereo panoramic vision for monitoring vehicle blind-spots
Author :
Matuszyk, Leanne ; Zelinsky, Alexander ; Nilsson, Lars ; Rilbe, Magnus
Author_Institution :
Res. Sch. of Inf. Sci. & Eng., Australian Nat. Univ., Canberra, ACT, Australia
Abstract :
This paper presents a stereo panoramic sensor as part of a driver assistance system to monitor driver blind-spots around vehicles. With our system we have generated panoramic disparity maps to reliably estimate range to objects in the surrounding environment. It was also proven that it is possible to apply the ν-disparity algorithm to panoramic images to successfully segment obstacles, even in the case of extremely noisy data. The stereo system has been evaluated using ground truth data, together with extensive field experiments.
Keywords :
driver information systems; image segmentation; image sensors; object detection; road vehicles; stereo image processing; driver assistance system; driver blind spots monitoring; ground truth data; noisy data; nu-disparity algorithm; obstacles segmentation; panoramic disparity maps; panoramic images; stereo panoramic sensor; stereo panoramic vision; stereo system; vehicle blind spots monitoring; Australia; Automotive engineering; Cameras; Image segmentation; Lenses; Mirrors; Monitoring; Robot vision systems; Sensor systems; Vehicle driving;
Conference_Titel :
Intelligent Vehicles Symposium, 2004 IEEE
Print_ISBN :
0-7803-8310-9
DOI :
10.1109/IVS.2004.1336351
