GYOLOv5 based x-ray security screening for dangerous goods detection

Hui-chao Hou; Gang Li; Meng-xia Sun; Kan He; Ling Zhang; Kang Ci

doi:10.1117/12.2689158

31 July 2023 GYOLOv5 based x-ray security screening for dangerous goods detection

Hui-chao Hou, Gang Li, Meng-xia Sun, Kan He, Ling Zhang, Kang Ci

Proceedings Volume 12747, Third International Conference on Optics and Image Processing (ICOIP 2023); 127471B (2023) https://doi.org/10.1117/12.2689158
Event: Third International Conference on Optics and Image Processing (ICOIP 2023), 2023, Hangzhou, China

Abstract

An X-ray security image detection model incorporating a multi-scale fusion module is proposed to address the problem of low accuracy in detecting threat objects in X-ray images against complex backgrounds. The model adds a multichannel fusion convolution block after the Neck layer to perform adaptive feature fusion and refinement on the input image, effectively improve the description of global information and boundary attributes of x-ray threat objects to improve the precision of detecting and identifying threat objects. SIoU is chosen to replace CIoU as the loss function of border regression, which redefines the penalty index and reduces the total degrees of freedom of loss to achieve the high accuracy localization. The model can effectively detect five different categories of dangerous goods on the Tianchi dataset, and the mAP value for dangerous goods detection is 92.7%, which is 2.1% higher than YOLOv5s, can satisfy the real-time recognition and detection requirements with high accuracy, good robustness and speed.

Citation Download Citation

Hui-chao Hou, Gang Li, Meng-xia Sun, Kan He, Ling Zhang, and Kang Ci "GYOLOv5 based x-ray security screening for dangerous goods detection", Proc. SPIE 12747, Third International Conference on Optics and Image Processing (ICOIP 2023), 127471B (31 July 2023); https://doi.org/10.1117/12.2689158

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