Abstract
This paper describes the results of a study aimed at inferring the surface structure of outdoor scenes from video data acquired by a vehicle-mounted TV camera. The method is based on a decomposition of each frame of the video sequence into a dense set of feature points. Through the application of a structure-from-motion algorithm, the 3D locations of the time-consistent features are estimated and sequentially updated as each new frame is processed. The resultant point-based 3D representation has been found to be reliable, accurate and rich enough to enable the surface structure of the viewed scene to be reconstructed. Examples of the surfaces extracted by this ‘bottom-up’ approach from a 12 second video sequence are presented here. The implications of these results for autonomous vehicle navigation are discussed.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
