Multi-view stereovision for translucent and semi-transparent objects
Master thesis
Problem and state of the art
Multi-view stereovision consists in recovering the three-dimensional shape (such as depth) of an object or a scene from several images of it. Nowaday, multi-view stereovision algorithms allow more or less to recover opaque objects; see for example Pons’ work or Jin and Soatto’s work. Despite the recent advances in opaque surface modeling, transparent surface modeling relatively has not received much attention. Also, successful methods aimed for opaque surface fail to deal with transparent surface. Only recently, some prospective techniques for modeling transparent surface have emerged. One of the most significative work in the field is perhaps the one of Fanany et al. which provides 3D shapes from a quite realistic senario.
Project
Contrary to the traditional multi-view stereovision algorithms which deal with opaque objects, we propose here to recover translucent and semi-transparent objects, as for example the following ones.
The constraints imposed by the scenario would be very free. One could control the lighting condition (photometric stereo) as well as the camera viewpoint (calibrated multiview reconstruction). It is for example, also possible to use background images as done in Fanany et al. paper.
Related work
"A neural network scheme for transparent surface modelling", M.I.Fanany, I.Kumazawa and K.Kobayashi
- Proceedings of the 3rd international conference on Computer graphics and interactive techniques in Australasia and South East Asia - 2005- Download PDF
"Stereo matching with reflections and translucency"
Y. Tsin, S. B. Kang and R. Szeliski - CVPR’03 -Download PDF
"3D-Reconstruction of Microscopic Translucent
Silicate-Based Marine and Freshwater-Organisms"
D. Schmitz, R. Herpers, D. Seibt and W. Heiden - Download PDF
"Three-dimensional shape measurement of transparent objects by triangulation approach" - D. Narita, M. Baba and K. Ohtani — SICE 2003 Annual Conference.
"How to grasp micro transparent object with integrated vision" — K. Ohara, M. Mizukawa, K. Ohba and T. Tanikawa — International Conference on Intelligent Robots and Systems, 2003.
Contact
The project will be carried out in the PERCEPTION group at INRIA Rhône-Alpes, under the supervision of Dr. Emmanuel Prados [ HTML],.
Related links
PERCEPTION Lab,
INRIA Grenoble homepage,
Start date: 1 November 2006
Contact person: Emmanuel PRADOS
Deadline: 1 January 2007
