Loading...
Please wait, while we are loading the content...
Similar Documents
2 - Gestion intelligente de capteurs et fusion multisensorielle pour la détection et le suivi d'obstacles sur route
| Content Provider | Semantic Scholar |
|---|---|
| Author | Trassoudaine, Laurent Checchin Alizon Collange Gallice |
| Copyright Year | 1996 |
| Abstract | In this article, we present a multisensorial solution for road obstacle detection and tracking . This solution is based on a mixe d camera/3D sensor mounted on the front of an experimental vehicle . The multisensor is described . The calibration step enables the matching of the heterogeneous data . Two capabalities of the senso r have been developped : the controlled perception making possible the acquisition of depth data in an area defined in the intensit y image; the visual servoing carrying out the focusing of the laser beam on a moving target detected in the intensity image . These two capabalities allow a Feedback control on the acquisition mode of the sensor according to the environment . ' The perception strategy is based on the selection of the best sensor for a given goal . The obstacle detection is based on th e segmentation and interpretation of depth data which are well suited in this context . However, the rate of acquisition of these dat a is too slow in order to extract the kinematic state of the obstacle . So, the tracking process is based on the collaboration betwee n intensity image processing which ensures the tracking itself and a 3D process which returns the obstacle model size to search in th e image. This algorithm of heterogeneous data fusion, associated with a Kalman filtering, permits to compute the state of obstacles . This work fits into the european project PROMETHEUS . Experimental results have been validated in real situation on the Prola b vehicle . |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://documents.irevues.inist.fr/bitstream/handle/2042/1951/002.PDF%20TEXTE.pdf?sequence=1 |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
