Development and Improvement of Optical Tracking Methods towards Registering the Deformations of 3D Non- Rigid Bodies in Real Time for Augmented Reality Applications
Year: 2015
Type: dissertation
Abstract: Augmented Reality (AR) is a technology that aims to embed virtual objects \nin the real world, showing to the user the set of objects (virtual and real) as \na single world. For that purpose, it is necessary to offer a perfect alignment \nbetween virtual and real objects, which increases the effectiveness of AR. The \nsolution to this problem is known as tracking. The object tracking consists in \ndetermining at any time the position and orientation of the camera relative to \nthe scene. Optical sensors are most commonly used to overcome the tracking \nproblem due to their low cost implementation. However, it is often difficult to \nprovide robustness, accuracy and low computational cost at the same time. \nThis thesis tackles the improvement and development of the main optical \ntracking techniques, primarily focused on detecting the deformations of the \nbodies. First, it has been achieved the tracking of rigid and non-rigid planar \nsurfaces through a monocular camera, and then, the object deformation estimation \nwith a more complex device as a RGB-D camera has been developed. \nSurface tracking systems such as those based on markers have the problem \nof not being able to handle occlusions. Thus, this thesis raises a new marker \ndesign that offers robustness against occlusions. Furthermore, in order to handle \nthe deformations of surfaces, a solution that recovers the camera pose and the \nnon-rigid surface simultaneously is proposed. Continuing with the deformation \nhandling, it has also developed a robust tracking system for reconstructing the \n3D shape of deformable objects using two different physical formulations. One \noffers a correct physical behaviour with a low computational cost, whereas the \nother achieves higher levels of accuracy at the expense of higher processing \ntime. \nIn addition, all the presented solutions have the common factor that all are \nexecuted in real time, which is a key property for a fluently visual feedback of \nan AR application.
Author: Leizea, Ibai
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W301688574