Tracking surfaces via texture-mapping: A boot-strapping approach
Mapping grey-level texture onto a surface model with known 3D structure enables tracking of the surface through a sequence of frames. The first frame position is assumed known. Since surface structure is known, grey-level texture can be mapped onto the 3D surface, and this texture is used to track the surface into the next frame; as more of the surface is revealed, more texture is mapped onto the surface which facilitates further tracking. Using this boot-strapping approach means that tracking and texture-mapping proceed simultaneously. Results are presented on a sequence of images in which a Champagne bottle is successfully tracked whilst rotating. At the end of the sequence the motion-path that the surface followed is known, and grey-level texture has been mapped onto the surface.
Unable to display preview. Download preview PDF.
- 1.Bray A J. Tracking objects Using image disparities. Proceedings of the 5th Alvey Vision Conference, Reading, 1989.Google Scholar
- 2.Bray A J. Recognising and Tracking Polyhedral Objects. PhD thesis, School of Cognitive and Computing Sciences, University of Sussex, UK, 1990.Google Scholar
- 3.Bray A J. Tracking curved objects by perspective inversion. Proceedings of the 2nd British Machine Vision Conference BMVC, 1991.Google Scholar
- 4.Stephens R S. Real-time 3D object tracking. Proceedings of the 5th Alvey Vision Conference, Reading, pages 85–90, 1989.Google Scholar
- 5.Harris C and Stennettt C. RAPID — a video rate object tracker. Proceedings of the 1st British Machine Vision Conference BMVC, pages 73–77, 1990.Google Scholar
- 6.Lowe D G. Stabilised solution for 3D model parameters. The First European Conference of Computer Vision, 1990.Google Scholar
- 7.Lowe D G. Fitting parameterised 3D models to images. IEEE Pattern Analysis and Machine Intelligence PAMI, 13(5):441–450, 1991.Google Scholar
- 8.North P R J. Reconstruction of visual appearance. Proceedings of the 1st British Machine Vision Conference BMVC, pages 205–210, 1990.Google Scholar
- 9.Lowe D G. Three-dimensional object recognition from two-dimensional images. Artificial Intelligence, 31(3), 1987.Google Scholar
- 10.Worrall A D, Baker K D, and Sullivan G D. Model based perspective inversion. Proceedings of the 4th Alvey Vision Conference, Manchester, pages 13–18, 1988.Google Scholar
- 11.Nishihara H K. Practical real-time imaging stereo matcher. Optical Engineering, 23(5):536–545, 1984.Google Scholar