EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

An Enhanced Pedestrian Detection Framework Based on the Fusion of Visible and Infrared Images Under Low-Light Conditions

Sally A.Abdulateef

Journal of Intelligent Decision Making and Information Science · 09.09.2026 · 3 (18s) · 1669-1680

DOI: 10.17762/jidmis.v3.5888 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Intelligent Decision Making and Information Science, online erschienen am 09.09.2026. Verfasst von Sally A.Abdulateef. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Due to the limited visual information that traditional visible-light cameras can gather, pedestrian recognition in low-light conditions continues to be one of the fundamental issues in computer vision. To overcome this problem, the present paper will provide a better pedestrian identification framework that uses a combination of visible and infrared images. The proposed approach initially improves contrast in HSV and L*a*b* color space of visible and infrared images before applying image fusion techniques. Two fusion methods are employed to produce informative fused images that do not lose thermal information or visual details: Gradient Transfer Fusion (GTF) and Convolutional Neural Network (CNN)-based fusion: The Faster R-CNN object detection model is then used to the improved fused images in order to detect pedestrians. The LLVIP and CVC-14 datasets were used for experimental evaluations. The obtained results indicate that the proposed improvement method significantly improves the low-light pedestrian identification. The GTF fusion strategy performed better than CNN-based fusion approach whose detection accuracy was 90% on the LLVIP dataset. The CVC-14 dataset improved as well, which indicated the effectiveness of the proposed preparation and fusion approach. These results suggest that combining multimodal image fusion with image enhancement can significantly increase the accuracy of pedestrian recognition in difficult lighting conditions.

Zitieren

APA

A.Abdulateef, S. (2026). An Enhanced Pedestrian Detection Framework Based on the Fusion of Visible and Infrared Images Under Low-Light Conditions. Journal of Intelligent Decision Making and Information Science, 3(18s), 1669-1680. https://doi.org/10.17762/jidmis.v3.5888

Vancouver

A.Abdulateef S. An Enhanced Pedestrian Detection Framework Based on the Fusion of Visible and Infrared Images Under Low-Light Conditions. Journal of Intelligent Decision Making and Information Science. 2026;3(18s):1669-1680. doi:10.17762/jidmis.v3.5888

BibTeX
@article{AAbdulateef2026,
  author = {A.Abdulateef, Sally},
  title = {An Enhanced Pedestrian Detection Framework Based on the Fusion of Visible and Infrared Images Under Low-Light Conditions},
  journal = {Journal of Intelligent Decision Making and Information Science},
  year = {2026},
  volume = {3},
  number = {18s},
  pages = {1669--1680},
  doi = {10.17762/jidmis.v3.5888}
}
Kostenlos starten →