ResearchSpace

Visual odometry for trailer off-tracking estimation

Show simple item record

dc.contributor.author De Saxe, Christopher
dc.contributor.author Cebon, D
dc.date.accessioned 2017-06-07T08:04:34Z
dc.date.available 2017-06-07T08:04:34Z
dc.date.issued 2016-11
dc.identifier.citation De Saxe, C. and Cebon, D. 2016. Visual odometry for trailer off-tracking estimation. 14th International Symposium on Heavy Vehicle Transport Technology (HVTT14), 15-18 November 2016, Rotorua, New Zealand en_US
dc.identifier.uri http://hvttconference.com/HVTT14/wp-content/uploads/2015/10/Conference_2016_Program_Reduced.pdf
dc.identifier.uri https://www.researchgate.net/publication/311038991_Visual_odometry_for_trailer_off-tracking_estimation
dc.identifier.uri http://hdl.handle.net/10204/9253
dc.description 14th International Symposium on Heavy Vehicle Transport Technology (HVTT14), 15-18 November 2016, Rotorua, New Zealand en_US
dc.description.abstract Application of articulated and Long Combination Vehicles (LCVs) in challenging off-highway applications is possible with the use of path-following trailer steering. This requires an accurate estimate of trailer off-tracking, but it has been shown that that existing methods for this are not applicable on roads with low friction or significant camber or grade. Here we propose an off-tracking measurement concept using stereo visual odometry which is applicable to off-highway environments. Simulation results demonstrate the theoretical accuracy of the system as well as the effects of camera placement and stereo baseline. Rear-mounted cameras are shown to yield the best precision, with RMS off-tracking measurement errors of 7–36 mm, while side-mounted cameras offer practical benefits such as scope for multiple-trailer configurations. Integration drift errors were shown to be bounded in time due to the relative nature of the off-tracking measurement. en_US
dc.language.iso en en_US
dc.relation.ispartofseries Worklist;17102
dc.subject Articulated HGVs en_US
dc.subject Off-tracking en_US
dc.subject Visual odometry en_US
dc.subject Stereo vision en_US
dc.subject Trailer steering en_US
dc.title Visual odometry for trailer off-tracking estimation en_US
dc.type Conference Presentation en_US
dc.identifier.apacitation De Saxe, C., & Cebon, D. (2016). Visual odometry for trailer off-tracking estimation. http://hdl.handle.net/10204/9253 en_ZA
dc.identifier.chicagocitation De Saxe, Christopher, and D Cebon. "Visual odometry for trailer off-tracking estimation." (2016): http://hdl.handle.net/10204/9253 en_ZA
dc.identifier.vancouvercitation De Saxe C, Cebon D, Visual odometry for trailer off-tracking estimation; 2016. http://hdl.handle.net/10204/9253 . en_ZA
dc.identifier.ris TY - Conference Presentation AU - De Saxe, Christopher AU - Cebon, D AB - Application of articulated and Long Combination Vehicles (LCVs) in challenging off-highway applications is possible with the use of path-following trailer steering. This requires an accurate estimate of trailer off-tracking, but it has been shown that that existing methods for this are not applicable on roads with low friction or significant camber or grade. Here we propose an off-tracking measurement concept using stereo visual odometry which is applicable to off-highway environments. Simulation results demonstrate the theoretical accuracy of the system as well as the effects of camera placement and stereo baseline. Rear-mounted cameras are shown to yield the best precision, with RMS off-tracking measurement errors of 7–36 mm, while side-mounted cameras offer practical benefits such as scope for multiple-trailer configurations. Integration drift errors were shown to be bounded in time due to the relative nature of the off-tracking measurement. DA - 2016-11 DB - ResearchSpace DP - CSIR KW - Articulated HGVs KW - Off-tracking KW - Visual odometry KW - Stereo vision KW - Trailer steering LK - https://researchspace.csir.co.za PY - 2016 T1 - Visual odometry for trailer off-tracking estimation TI - Visual odometry for trailer off-tracking estimation UR - http://hdl.handle.net/10204/9253 ER - en_ZA


Files in this item

This item appears in the following Collection(s)

Show simple item record