Automatic Camera Pose Estimation by Key-Point Matching of Reference Objects

Conference paper (2023)

Authors

Jinchen Zeng Student
R.M. Butler Medical Instruments & Bio-Inspired Technology - Mechanical, Maritime and Materials Engineering
J.J. van den Dobbelsteen Medical Instruments & Bio-Inspired Technology - Mechanical, Maritime and Materials Engineering
B.H.W. Hendriks Medical Instruments & Bio-Inspired Technology - Mechanical, Maritime and Materials Engineering , Philips Research Laboratories
M. van der Elst Reinier de Graaf Gasthuis, Support Biomechanical Engineering
J.H.G. Dauwels Signal Processing Systems - EEMCS
Research Group
Medical Instruments & Bio-Inspired Technology (Mechanical, Maritime and Materials Engineering) (TU Delft)
Copyright: © 2023 Jinchen Zeng, R.M. Butler, J.J. van den Dobbelsteen, B.H.W. Hendriks, M. van der Elst, J.H.G. Dauwels
DOI: https://doi.org/10.1109/ICASSP49357.2023.10095197
Camera calibration Camera pose estimation Perspective-n-Point 3D geometry
More Info
expand_more

Published Date

2023

Language

English

Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Faculty

Mechanical, Maritime and Materials Engineering

Department

Biomechanical Engineering

Research Group

Medical Instruments & Bio-Inspired Technology

Abstract

In this paper, we aim to design an automatic camera pose estimation pipeline for clinical spaces such as catheterization laboratories. Our proposed pipeline exploits Scaled-YOLOv4 to detect fixed objects. We adopt the self-supervised key-point detector SuperPoint in combination with SuperGlue, a keypoint matching technique based on graph neural networks. Thus, we match key-points on input images with annotated reference points. Reference points are chosen on fixed objects in the scene, such as corners of door posts or windows. The point-correspondences between the image coordinates and the 3D coordinates are applied to the Perspective-n-Point algorithm to estimate the pose of each camera. Compared with other camera pose estimation methods, the proposed pipeline does not require the construction of 3D point-cloud model of the scene or placing a polyhedron object in the scene before each required calibration. Using videos from real procedures, we show that the pipeline can estimate the camera pose with high accuracy.