Skip to main navigation Skip to search Skip to main content

Refined Attitude Estimation of Ships in Photographs via Matching Images Rendered from 3D Models

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

Abstract

In this paper, an iterative procedure is proposed to estimate the attitude of objects in photographs via matching images rendered from three-dimensional (3D) models, which have six degrees of freedom (6-DOF) motions such as ships on sea surface. The perspective projection matrix from 3D to two-dimensional (2D) coordinates is obtained using the principle of camera imaging process in computer graphics. Seven parameters are first coarsely estimated using the least square method, including the field of view (FOV) of camera and the 6-DOF attitude of an object. An iteration process is then implemented to refine the attitude estimates within specified ranges. The 6-DOF attitude of an oil tanker in photographs is estimated as examples. And the perfect results demonstrate the usefulness of the proposed procedure.

Original languageEnglish
Title of host publication2017 Sensor Signal Processing for Defence Conference, SSPD 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1-5
Number of pages5
ISBN (Electronic)9781538616635
DOIs
StatePublished - 20 Dec 2017
Event7th Conference of the Sensor Signal Processing for Defence, SSPD 2017 - London, United Kingdom
Duration: 6 Dec 20177 Dec 2017

Publication series

Name2017 Sensor Signal Processing for Defence Conference, SSPD 2017
Volume2017-January

Conference

Conference7th Conference of the Sensor Signal Processing for Defence, SSPD 2017
Country/TerritoryUnited Kingdom
CityLondon
Period6/12/177/12/17

Keywords

  • Attitude estimation
  • Iterative procedure
  • Least square method
  • Perspective projection
  • Three-dimensional model matching

Fingerprint

Dive into the research topics of 'Refined Attitude Estimation of Ships in Photographs via Matching Images Rendered from 3D Models'. Together they form a unique fingerprint.

Cite this