An international conference connecting people
in CAD research, education and business
Bookmark and Share
Copyright (C) CAD Solutions, LLC. All rights reserved.
Proceedings of CAD'14, 2014, 206-208
Non-Rigid 3D Object Pose Normalization based on Core Segment Symmetrical Features

Konstantinos Sfikas, Theoharis Theoharis, University of Athens, Greece
Ioannis Pratikakis, Democritus University of Thrace, Greece

Abstract. Pose normalization is an important preprocessing step in a variety of computer graphics applications. Visualization, broken fragment reconstruction, biometrics and 3D object retrieval are only a few examples of applications that benefit from a pose normalization procedure. Research in pose normalization has given some very efficient methodologies, mainly based on variations of principal component analysis (PCA), like the Continuous PCA and the Normal PCA methods. Another category of methods that effectively deal with the problem is the one exploiting characteristics of symmetry found in a plethora of 3D objects. Pose normalization has mainly focused on rigid 3D objects whereas there has been little research on non-rigid 3D objects. In this paper, we propose a method that achieves pose normalization of non-rigid 3D objects by taking into account the core segment of the object. Major obstacle in the pose normalization of non-rigid 3D objects is the variety of deformations that they exhibit. More specifically, while most segments of a non-rigid 3D object remain unchanged, deformations occurring in articulating areas and protrusions alter the shape of the objects significantly. The proposed method aims at the extraction of a segment of the 3D object with the minimal deformation which can be used for pose normalization.

Keywords. Pose normalization, non-rigid 3D objects

DOI: 10.14733/cadconfP.2014.206-208