Pose estimation and 3D environment reconstruction using less reliable depth data

Sungjin Jo, Hyung Gi Jo, Hae Min Cho, Euntai Kim

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Scopus citations

Abstract

Pose estimation and 3D reconstruction of environment are essential technics in robotics and computer vision. In this paper we present a method for camera tracking and 3D reconstruction of static environments, using a ToF sensor which provides less reliable depth information. Based on a primary camera pose, we eliminate outlier in distance measurements. Subsequently, we estimate camera pose again using only inlier data. A voxel grid map is updated by integrating depth measurement with a truncated signed distance function. It is represented as 3D environment reconstruction. Our method is an attractive extending of the pose estimation in outdoor environment. In outdoor environment, 3D range cameras cannot measure the distance or they provide inaccurate distance measurement. The experiments were carried out both in indoor and outdoor and we analyze the results of the proposed methods which use a ToF camera in comparison with a previous approach.

Original languageEnglish
Title of host publicationAIM 2015 - 2015 IEEE/ASME International Conference on Advanced Intelligent Mechatronics
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages359-364
Number of pages6
ISBN (Electronic)9781467391078
DOIs
StatePublished - 25 Aug 2015
EventIEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM 2015 - Busan, Korea, Republic of
Duration: 7 Jul 201511 Jul 2015

Publication series

NameIEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM
Volume2015-August

Conference

ConferenceIEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM 2015
Country/TerritoryKorea, Republic of
CityBusan
Period7/07/1511/07/15

Fingerprint

Dive into the research topics of 'Pose estimation and 3D environment reconstruction using less reliable depth data'. Together they form a unique fingerprint.

Cite this