This paper presents an engine for texturing 3D models from real-world images that implements a novel approach to simplify the user's interactions when registering images to geometry. Indeed, the number of user's interactions is significantly reduced and not requiring high precision; in addition, those interactions are simple as it consists only in dragging operations controlled via real time feedbacks. The key idea is to take advantage of a three-dimensional orientation sensor attached to the camera so as to make the object pose estimation simpler. Geometric computation is introduced to implement the idea. Furthermore, with the aim of robustly refining the pose, an adaptation of an existing tracking method is presented. A set of experiments demonstrating the efficiency and practicality of this approach has been conducted.
Published in:
Computer and Information Technology, 2009. CIT '09. Ninth IEEE International Conference on
(Volume:1
)
Date of Conference: 11-14 Oct. 2009