SLAM using both points and planes for hand-held 3D sensors

Yuichi Taguchi, Yong Dian Jian, Srikumar Ramalingam, Chen Feng

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We present a simultaneous localization and mapping (SLAM) algorithm for a hand-held 3D sensor that uses both points and planes as primitives. Our algorithm uses any combination of three point/plane primitives (3 planes, 2 planes and 1 point, 1 plane and 2 points, and 3 points) in a RANSAC framework to efficiently compute the sensor pose. As the number of planes is significantly smaller than the number of points in typical 3D scenes, our RANSAC algorithm prefers primitive combinations involving more planes than points. In contrast to existing approaches that mainly use points for registration, our algorithm has the following advantages: (1) it enables faster correspondence search and registration due to the smaller number of plane primitives; (2) it produces plane-based 3D models that are more compact than point-based ones; and (3) being a global registration algorithm, our approach does not suffer from local minima or any initialization problems. Our experiments demonstrate real-time, interactive 3D reconstruction of office spaces using a hand-held Kinect sensor.

Original languageEnglish (US)
Title of host publicationISMAR 2012 - 11th IEEE International Symposium on Mixed and Augmented Reality 2012, Science and Technology Papers
Pages321-322
Number of pages2
DOIs
StatePublished - Dec 1 2012
Event11th IEEE and ACM International Symposium on Mixed and Augmented Reality, ISMAR 2012 - Atlanta, GA, United States
Duration: Nov 5 2012Nov 8 2012

Other

Other11th IEEE and ACM International Symposium on Mixed and Augmented Reality, ISMAR 2012
CountryUnited States
CityAtlanta, GA
Period11/5/1211/8/12

Fingerprint

Sensors
Experiments

Keywords

  • I.4.8 [Image Processing and Computer Vision]: Scene Analysis - Range Data
  • Tracking

ASJC Scopus subject areas

  • Artificial Intelligence

Cite this

Taguchi, Y., Jian, Y. D., Ramalingam, S., & Feng, C. (2012). SLAM using both points and planes for hand-held 3D sensors. In ISMAR 2012 - 11th IEEE International Symposium on Mixed and Augmented Reality 2012, Science and Technology Papers (pp. 321-322). [6402594] https://doi.org/10.1109/ISMAR.2012.6402594

SLAM using both points and planes for hand-held 3D sensors. / Taguchi, Yuichi; Jian, Yong Dian; Ramalingam, Srikumar; Feng, Chen.

ISMAR 2012 - 11th IEEE International Symposium on Mixed and Augmented Reality 2012, Science and Technology Papers. 2012. p. 321-322 6402594.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Taguchi, Y, Jian, YD, Ramalingam, S & Feng, C 2012, SLAM using both points and planes for hand-held 3D sensors. in ISMAR 2012 - 11th IEEE International Symposium on Mixed and Augmented Reality 2012, Science and Technology Papers., 6402594, pp. 321-322, 11th IEEE and ACM International Symposium on Mixed and Augmented Reality, ISMAR 2012, Atlanta, GA, United States, 11/5/12. https://doi.org/10.1109/ISMAR.2012.6402594
Taguchi Y, Jian YD, Ramalingam S, Feng C. SLAM using both points and planes for hand-held 3D sensors. In ISMAR 2012 - 11th IEEE International Symposium on Mixed and Augmented Reality 2012, Science and Technology Papers. 2012. p. 321-322. 6402594 https://doi.org/10.1109/ISMAR.2012.6402594
Taguchi, Yuichi ; Jian, Yong Dian ; Ramalingam, Srikumar ; Feng, Chen. / SLAM using both points and planes for hand-held 3D sensors. ISMAR 2012 - 11th IEEE International Symposium on Mixed and Augmented Reality 2012, Science and Technology Papers. 2012. pp. 321-322
@inproceedings{7dbad37adea447f9bde25ff2d9933fa9,
title = "SLAM using both points and planes for hand-held 3D sensors",
abstract = "We present a simultaneous localization and mapping (SLAM) algorithm for a hand-held 3D sensor that uses both points and planes as primitives. Our algorithm uses any combination of three point/plane primitives (3 planes, 2 planes and 1 point, 1 plane and 2 points, and 3 points) in a RANSAC framework to efficiently compute the sensor pose. As the number of planes is significantly smaller than the number of points in typical 3D scenes, our RANSAC algorithm prefers primitive combinations involving more planes than points. In contrast to existing approaches that mainly use points for registration, our algorithm has the following advantages: (1) it enables faster correspondence search and registration due to the smaller number of plane primitives; (2) it produces plane-based 3D models that are more compact than point-based ones; and (3) being a global registration algorithm, our approach does not suffer from local minima or any initialization problems. Our experiments demonstrate real-time, interactive 3D reconstruction of office spaces using a hand-held Kinect sensor.",
keywords = "I.4.8 [Image Processing and Computer Vision]: Scene Analysis - Range Data, Tracking",
author = "Yuichi Taguchi and Jian, {Yong Dian} and Srikumar Ramalingam and Chen Feng",
year = "2012",
month = "12",
day = "1",
doi = "10.1109/ISMAR.2012.6402594",
language = "English (US)",
isbn = "9781467346603",
pages = "321--322",
booktitle = "ISMAR 2012 - 11th IEEE International Symposium on Mixed and Augmented Reality 2012, Science and Technology Papers",

}

TY - GEN

T1 - SLAM using both points and planes for hand-held 3D sensors

AU - Taguchi, Yuichi

AU - Jian, Yong Dian

AU - Ramalingam, Srikumar

AU - Feng, Chen

PY - 2012/12/1

Y1 - 2012/12/1

N2 - We present a simultaneous localization and mapping (SLAM) algorithm for a hand-held 3D sensor that uses both points and planes as primitives. Our algorithm uses any combination of three point/plane primitives (3 planes, 2 planes and 1 point, 1 plane and 2 points, and 3 points) in a RANSAC framework to efficiently compute the sensor pose. As the number of planes is significantly smaller than the number of points in typical 3D scenes, our RANSAC algorithm prefers primitive combinations involving more planes than points. In contrast to existing approaches that mainly use points for registration, our algorithm has the following advantages: (1) it enables faster correspondence search and registration due to the smaller number of plane primitives; (2) it produces plane-based 3D models that are more compact than point-based ones; and (3) being a global registration algorithm, our approach does not suffer from local minima or any initialization problems. Our experiments demonstrate real-time, interactive 3D reconstruction of office spaces using a hand-held Kinect sensor.

AB - We present a simultaneous localization and mapping (SLAM) algorithm for a hand-held 3D sensor that uses both points and planes as primitives. Our algorithm uses any combination of three point/plane primitives (3 planes, 2 planes and 1 point, 1 plane and 2 points, and 3 points) in a RANSAC framework to efficiently compute the sensor pose. As the number of planes is significantly smaller than the number of points in typical 3D scenes, our RANSAC algorithm prefers primitive combinations involving more planes than points. In contrast to existing approaches that mainly use points for registration, our algorithm has the following advantages: (1) it enables faster correspondence search and registration due to the smaller number of plane primitives; (2) it produces plane-based 3D models that are more compact than point-based ones; and (3) being a global registration algorithm, our approach does not suffer from local minima or any initialization problems. Our experiments demonstrate real-time, interactive 3D reconstruction of office spaces using a hand-held Kinect sensor.

KW - I.4.8 [Image Processing and Computer Vision]: Scene Analysis - Range Data

KW - Tracking

UR - http://www.scopus.com/inward/record.url?scp=84873553704&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84873553704&partnerID=8YFLogxK

U2 - 10.1109/ISMAR.2012.6402594

DO - 10.1109/ISMAR.2012.6402594

M3 - Conference contribution

AN - SCOPUS:84873553704

SN - 9781467346603

SP - 321

EP - 322

BT - ISMAR 2012 - 11th IEEE International Symposium on Mixed and Augmented Reality 2012, Science and Technology Papers

ER -