GL-YOLO: An Enhanced YOLOv11-Based Model for Steel Surface Defect Detection.

Saved in:

Bibliographic Details
Title:	GL-YOLO: An Enhanced YOLOv11-Based Model for Steel Surface Defect Detection.
Authors:	Wu, Fanghan¹ 1514705165@qq.com, Zhang, Yujun¹ 1997zyj@163.com
Source:	Engineering Letters. May2026, Vol. 34 Issue 5, p1941-1952. 12p.
Subjects:	Multiscale modeling, Object recognition (Computer vision), Manufacturing process automation, Automation
Abstract:	With the rapid advancement of industrial automation and intelligent manufacturing, steel surface defect detection increasingly demands both high detection accuracy and real-time performance. To address these requirements, a high-precision steel surface defect detection model, termed GL-YOLO, is proposed based on YOLOv11n, aiming to improve defect recognition under complex background conditions. The proposed improvements focus on three aspects. First, a Global-Local C3 Block (GLC3 Block) is incorporated into the YOLOv11n backbone, where a controllable global-local receptive field mechanism is employed to adaptively model multi-scale features, thereby enhancing the representation of diverse defect textures. Second, a CA-Mix Block is introduced at the P3 feature extraction stage to strengthen spatial and channel-wise feature interactions. Third, a P2 Small Object Head is added to improve the detection of small-scale and weak-texture defects. Experimental results on the NEU-DET steel surface defect dataset demonstrate that GL-YOLO achieves a 4.6% improvement in mean average precision (mAP@50) compared with the baseline YOLOv11n. These results indicate that the proposed GL-YOLO effectively enhances detection accuracy and consistently outperforms the baseline model. [ABSTRACT FROM AUTHOR]
	Copyright of Engineering Letters is the property of International Association of Engineers (IAENG) and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database:	Engineering Source

Description
Abstract:	With the rapid advancement of industrial automation and intelligent manufacturing, steel surface defect detection increasingly demands both high detection accuracy and real-time performance. To address these requirements, a high-precision steel surface defect detection model, termed GL-YOLO, is proposed based on YOLOv11n, aiming to improve defect recognition under complex background conditions. The proposed improvements focus on three aspects. First, a Global-Local C3 Block (GLC3 Block) is incorporated into the YOLOv11n backbone, where a controllable global-local receptive field mechanism is employed to adaptively model multi-scale features, thereby enhancing the representation of diverse defect textures. Second, a CA-Mix Block is introduced at the P3 feature extraction stage to strengthen spatial and channel-wise feature interactions. Third, a P2 Small Object Head is added to improve the detection of small-scale and weak-texture defects. Experimental results on the NEU-DET steel surface defect dataset demonstrate that GL-YOLO achieves a 4.6% improvement in mean average precision (mAP@50) compared with the baseline YOLOv11n. These results indicate that the proposed GL-YOLO effectively enhances detection accuracy and consistently outperforms the baseline model. [ABSTRACT FROM AUTHOR]
ISSN:	1816093X