Implementation and Accuracy Evaluation of Fixed Camera-Based Object Positioning System Employing CNN-Detector

Tero Partanen, Philipp Muller, Jussi Collin, Jani Björklund

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

9 Downloads (Pure)

Abstract

Today, different positioning applications such as location-based services and autonomous navigation are requiring more and more precision. Especially fully autonomous navigation requires accurate positioning solution, not only for the vehicle but also for the surrounding objects. Thus, many new positioning techniques, algorithms and fusion schemes have been developed. One essential technique is visual positioning. Thanks to intensive research in neural networks and deep learning, Convolutional Neural Network-based (CNN) object detectors have evolved greatly in recent years. This paper proposes a widely deployable scheme of fixed camera-based (e.g. surveillance camera) object positioning utilizing the CNN-detector. The accuracy of the implemented positioning solution is evaluated with precise Real-Time Kinematic (RTK) satellite positioning receiver. The implemented system can be used in indoors and outdoors, and it can estimate simultaneously positions from multiple camera views for multiple objects in real-Time. When positioning a person, the measured mean positioning error was 10.7-15.6 cm with a simple bias correction and a standard deviation was 6.7-8.7 cm. Thus, the accuracy is excellent and would be sufficient to wide variety of applications.

Original languageEnglish
Title of host publicationProceedings of the 2021 9th European Workshop on Visual Information Processing, EUVIP 2021
EditorsA. Beghdadi, F. Alaya Cheikh, J.M.R.S. Tavares, A. Mokraoui, G. Valenzise, L. Oudre, M.A. Qureshi
PublisherIEEE
Number of pages6
ISBN (Electronic)9781665432306
ISBN (Print)9781665432313
DOIs
Publication statusPublished - 2021
Publication typeA4 Article in a conference publication
EventEuropean Workshop on Visual Information Processing - Paris, France
Duration: 23 Jun 202125 Jun 2021

Publication series

NameEuropean Workshop on Visual Information Processing
ISSN (Print)2164-974X
ISSN (Electronic)2471-8963

Conference

ConferenceEuropean Workshop on Visual Information Processing
Country/TerritoryFrance
CityParis
Period23/06/2125/06/21

Keywords

  • Camera-based positioning
  • convolutional neural networks (CNN)
  • object detection
  • RTK GNSS
  • visual localization

Publication forum classification

  • Publication forum level 1

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Information Systems
  • Signal Processing

Fingerprint

Dive into the research topics of 'Implementation and Accuracy Evaluation of Fixed Camera-Based Object Positioning System Employing CNN-Detector'. Together they form a unique fingerprint.

Cite this