Inferring 3D dimensions of flat objects from a single 2D image

Sreedevi Sreekaladevi, Wai Lok Woo, Satnam Dlay, Peter Gore, Garth Johnson, David Swailes

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

4 Citations (Scopus)

Abstract

In this work, a method to infer 3D dimensions of flat objects from a single image is presented. The idea is based on the computation of position of a reference object lying in space. A sheet of A4 paper is used as the reference object and the camera is calibrated using the four corners of the paper. Once the position of reference paper is known, further measurements can be taken on the plane in which it lies. No distortion correction model has been implemented yet. This method has been tested on various three dimensional scenarios and it has been found that the accuracy of measurements inferred lie within a tolerance of 25 mm.

Original languageEnglish
Title of host publicationProceedings - 2013 IEEE 9th International Conference on Intelligent Computer Communication and Processing, ICCP 2013
PublisherIEEE
Pages123-126
Number of pages4
ISBN (Electronic)9781479914944
ISBN (Print)9781479914937
DOIs
Publication statusPublished - 24 Oct 2013
Event2013 IEEE 9th International Conference on Intelligent Computer Communication and Processing, ICCP 2013 - Cluj-Napoca, Romania
Duration: 5 Sept 20137 Sept 2013

Conference

Conference2013 IEEE 9th International Conference on Intelligent Computer Communication and Processing, ICCP 2013
Country/TerritoryRomania
CityCluj-Napoca
Period5/09/137/09/13

Keywords

  • Cameras
  • Calibration
  • Accuracy
  • Three-dimensional displays
  • Mathematical model
  • Measurement uncertainty
  • Equations

Fingerprint

Dive into the research topics of 'Inferring 3D dimensions of flat objects from a single 2D image'. Together they form a unique fingerprint.

Cite this