Infrared small-scale pedestrian detection based on YOLO

Hongchang Zhang; Kang Yang; Jiali Hu

doi:10.1117/12.3033707

13 June 2024 Infrared small-scale pedestrian detection based on YOLO

Hongchang Zhang, Kang Yang, Jiali Hu

Proceedings Volume 13180, International Conference on Image, Signal Processing, and Pattern Recognition (ISPP 2024); 131806H (2024) https://doi.org/10.1117/12.3033707
Event: International Conference on Image, Signal Processing, and Pattern Recognition (ISPP 2024), 2024, Guangzhou, China

Abstract

Infrared small-scale pedestrian detection is particularly important at night and under low-light conditions, but it also faces great challenges, such as the small size of pedestrians and unclear target features. In order to solve these problems, this paper introduces two attention mechanisms based on the YOLOv5s model: channel attention mechanism (CA) and efficient channel attention mechanism (ECA) to enhance the model's ability to identify infrared pedestrian targets. By integrating these attention mechanisms in the YOLOv5s model, this paper aims to improve the accuracy and robustness of infrared pedestrian detection. In addition, this study also specially created an infrared pedestrian image dataset to train and test the proposed optimization model, and verified the effectiveness of the model through ablation experiments. Experimental results show that compared with the original YOLOv5s model, the optimized model achieved an improvement of 1.9 percentage points in average accuracy (AP), reaching 94.1%, which proves the effectiveness of the proposed method and its performance in infrared pedestrian detection tasks adaptability.

(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Hongchang Zhang, Kang Yang, and Jiali Hu "Infrared small-scale pedestrian detection based on YOLO", Proc. SPIE 13180, International Conference on Image, Signal Processing, and Pattern Recognition (ISPP 2024), 131806H (13 June 2024); https://doi.org/10.1117/12.3033707

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available