Enhanced Small-Object Detection on PCBs: YOLOv8-based Resolution Upscaling with Additional Fusion Layer Implementation.

Saved in:

Bibliographic Details
Title:	Enhanced Small-Object Detection on PCBs: YOLOv8-based Resolution Upscaling with Additional Fusion Layer Implementation.
Authors:	Tseng, Shih-Hsien¹ shtseng@mail.ntust.edu.tw, Nguyen, Thi Thu Ha¹, Duong, Thi Ha Trang¹
Source:	Journal of Imaging Science & Technology. Mar/Apr2026, Vol. 70 Issue 2, p1-13. 13p.
Subjects:	Printed circuits industry, Detection algorithms, Spatial resolution, Real-time computing, Artificial neural networks, Object recognition (Computer vision)
Abstract:	The determined drive toward miniaturization in electronics has led to increasingly complex printed circuit board (PCB) designs, posing significant challenges for object detection and inspection processes. This research introduces an innovative method to improve the detection of small objects on PCBs by integrating advanced multiscale layer fusion techniques with the YOLOv8 object detection framework. Leveraging the capabilities of YOLOv8, the proposed methodology addresses the limitations imposed by low-resolution imaging systems, thereby enhancing the reliability and accuracy of small-object detection. The effectiveness of the proposed approach is assessed through experimentation and validation, showcasing its ability to detect small components and defects on PCBs. The results indicate superior performance compared to existing methods, with a mean Average Precision (mAP@0.5) of 99.30% and an inference speed of 161 frames per second (FPS). This high FPS and high accuracy facilitate real-time processing, making the model suitable for deployment in time-sensitive industrial environments. [ABSTRACT FROM AUTHOR]
	Copyright of Journal of Imaging Science & Technology is the property of International Society for Imaging Science & Technology 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:	The determined drive toward miniaturization in electronics has led to increasingly complex printed circuit board (PCB) designs, posing significant challenges for object detection and inspection processes. This research introduces an innovative method to improve the detection of small objects on PCBs by integrating advanced multiscale layer fusion techniques with the YOLOv8 object detection framework. Leveraging the capabilities of YOLOv8, the proposed methodology addresses the limitations imposed by low-resolution imaging systems, thereby enhancing the reliability and accuracy of small-object detection. The effectiveness of the proposed approach is assessed through experimentation and validation, showcasing its ability to detect small components and defects on PCBs. The results indicate superior performance compared to existing methods, with a mean Average Precision (mAP@0.5) of 99.30% and an inference speed of 161 frames per second (FPS). This high FPS and high accuracy facilitate real-time processing, making the model suitable for deployment in time-sensitive industrial environments. [ABSTRACT FROM AUTHOR]
ISSN:	10623701
DOI:	10.2352/J.ImagingSci.Technol.2026.70.2.020507