EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Historical Blurry Video-Based Face Recognition

Lujun Zhai iD, Suxia Cui, Yonghui Wang, Song Wang, Jun Zhou, Greg Wilsbacher

Journal of Imaging · 20.09.2024 · 10 (9) · 236

DOI: 10.3390/jimaging10090236 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Imaging, online erschienen am 20.09.2024. Verfasst von Lujun Zhai und 5 weiteren Autor:innen. Beteiligte Einrichtung: Electrical and Computer Engineering Department, Prairie View A&M University, Prairie View, TX 77446, USA. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 61 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: NSF, US Department of Education.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Face recognition is a widely used computer vision, which plays an increasingly important role in user authentication systems, security systems, and consumer electronics. The models for most current applications are based on high-definition digital cameras. In this paper, we focus on digital images derived from historical motion picture films. Historical motion picture films often have poorer resolution than modern digital imagery, making face detection a more challenging task. To approach this problem, we first propose a trunk–branch concatenated multi-task cascaded convolutional neural network (TB-MTCNN), which efficiently extracts facial features from blurry historical films by combining the trunk with branch networks and employing various sizes of kernels to enrich the multi-scale receptive field. Next, we build a deep neural network-integrated object-tracking algorithm to compensate for failed recognition over one or more video frames. The framework combines simple online and real-time tracking with deep data association (Deep SORT), and TB-MTCNN with the residual neural network (ResNet) model. Finally, a state-of-the-art image restoration method is employed to reduce the effect of noise and blurriness. The experimental results show that our proposed joint face recognition and tracking network can significantly reduce missed recognition in historical motion picture film frames.

Schlagwörter

Stichwörter der Autor:innen

face detectionface recognitionface trackinghistorical blurry video

Autor:innen und Institutionen

  • Electrical and Computer Engineering Department, Prairie View A&M University, Prairie View, TX 77446, USA.
  • Computer Science Department, Prairie View A&M University, Prairie View, TX 77446, USA.
  • Computer Science and Engineering Department, University of South Carolina, Columbia, SC 29425, USA.
  • Moving Image Research Collections, University Libraries, University of South Carolina, Columbia, SC 29425, USA.

ORCID-Profile: Lujun Zhai

Förderung

  • NSF
  • US Department of Education

Zitieren

APA

Zhai, L., Cui, S., Wang, Y., Wang, S., Zhou, J., & Wilsbacher, G. (2024). Historical Blurry Video-Based Face Recognition. Journal of Imaging, 10(9), 236. https://doi.org/10.3390/jimaging10090236

Vancouver

Zhai L, Cui S, Wang Y, Wang S, Zhou J, Wilsbacher G. Historical Blurry Video-Based Face Recognition. Journal of Imaging. 2024;10(9):236. doi:10.3390/jimaging10090236

BibTeX
@article{Zhai2024,
  author = {Zhai, Lujun and Cui, Suxia and Wang, Yonghui and Wang, Song and Zhou, Jun and Wilsbacher, Greg},
  title = {Historical Blurry Video-Based Face Recognition},
  journal = {Journal of Imaging},
  year = {2024},
  volume = {10},
  number = {9},
  pages = {236},
  doi = {10.3390/jimaging10090236}
}
Kostenlos starten →