Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Three Dimensional Posed Face Recognition with an Improved Iterative Closest Point Method

Three Dimensional Posed Face Recognition with an Improved Iterative Closest Point Method Dealing with different head poses is one of the most challenging issues in the area of face recognition. Recently, 3D images have been used for this purpose as they can gather more information from the head area. Kinect was used for capturing 3D images in our research. Iterative Closest Point (ICP) algorithm has been used in many researches to align a rotated pointcloud with its corresponding reference. However it has many variables that can improve its performance. So an improved version of ICP has been introduced in our research and its performance in terms of accuracy and speed has been evaluated. While it can have many applications, we have used it for increasing the performance of posed face recognition. We applied our proposed algorithm on a local database and concluded that it can significantly improve the recognition rate of 3D posed face recognition compared with using original raw posed image. Results of executing the proposed algorithm on a public database also indicate an improvement with respect to other recently proposed methods. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png 3D Research Springer Journals

Three Dimensional Posed Face Recognition with an Improved Iterative Closest Point Method

3D Research , Volume 10 (4) – Jul 13, 2019

Loading next page...
 
/lp/springer-journals/three-dimensional-posed-face-recognition-with-an-improved-iterative-M5Bood3sNb
Publisher
Springer Journals
Copyright
Copyright © 2019 by 3D Display Research Center, Kwangwoon University and Springer-Verlag GmbH Germany, part of Springer Nature
Subject
Engineering; Signal,Image and Speech Processing; Computer Imaging, Vision, Pattern Recognition and Graphics; Optics, Lasers, Photonics, Optical Devices
eISSN
2092-6731
DOI
10.1007/s13319-019-0232-0
Publisher site
See Article on Publisher Site

Abstract

Dealing with different head poses is one of the most challenging issues in the area of face recognition. Recently, 3D images have been used for this purpose as they can gather more information from the head area. Kinect was used for capturing 3D images in our research. Iterative Closest Point (ICP) algorithm has been used in many researches to align a rotated pointcloud with its corresponding reference. However it has many variables that can improve its performance. So an improved version of ICP has been introduced in our research and its performance in terms of accuracy and speed has been evaluated. While it can have many applications, we have used it for increasing the performance of posed face recognition. We applied our proposed algorithm on a local database and concluded that it can significantly improve the recognition rate of 3D posed face recognition compared with using original raw posed image. Results of executing the proposed algorithm on a public database also indicate an improvement with respect to other recently proposed methods.

Journal

3D ResearchSpringer Journals

Published: Jul 13, 2019

References