The problem of updating the global position of an autonomous vehicle is considered. An iterative procedure is proposed to fit a map to a set of noisy measurements. The procedure is inspired by a non-parametric procedure for probability density function mode searching. We show how this could be used to solve the global position update problem in an efficient and robust way. The simple procedure combines the a priori knowledge from the known map of the environment, the extracted information from calibrated camera images and the information from the vehicle wheels rotation measurements. Because it is very simple the procedure is appropriate to be performed in real-time.
|Title of host publication||Proceedings of the 8th Mechatronics Forum International Conference, Mechatronics 2000|
|Subtitle of host publication||University of Twente, Netherlands, 24-26 June 2002|
|Editors||J. van Amerongen, B. Jonker, P.P.L Regtien, S. Stramigioli|
|Place of Publication||Enschede, Netherlands|
|Publisher||Drebbel Institute for Mechatronics|
|Number of pages||9|
|Publication status||Published - Jun 2002|
|Event||8th Mechatronics Forum International Conference, MECHATRONICS 2002 - Enschede, Netherlands|
Duration: 24 Jun 2002 → 26 Jun 2002
Conference number: 8
|Conference||8th Mechatronics Forum International Conference, MECHATRONICS 2002|
|Abbreviated title||Mechatronics 2002|
|Period||24/06/02 → 26/06/02|
Zivkovic, Z., Schoute, A., & van der Heijden, F. (2002). Combining A Priori Knowledge and Sensor Information for Updating the Global Position of an Autonomous Vehicle. In J. van Amerongen, B. Jonker, P. P. L. Regtien, & S. Stramigioli (Eds.), Proceedings of the 8th Mechatronics Forum International Conference, Mechatronics 2000: University of Twente, Netherlands, 24-26 June 2002 (pp. 643-651). Enschede, Netherlands: Drebbel Institute for Mechatronics.