Article: PCBA defect classification in real world with out-of-distribution detection Journal: International Journal of System Control and Information Processing (IJSCIP) 2025 Vol.4 No.3 pp.187 - 203 Abstract: Inadequate chip solder joints can significantly impact the overall quality of finalised Printed Circuit Board (PCB) products. Detecting all solder joint defects in real time and adaptively during the actual production process poses a formidable challenge due to the diverse nature of these defects and the limited availability of anomaly data. In contrast to the conventional Printed Circuit Board Assembly (PCBA) defect classification task that is limited to pre-labelled defects, we propose adaptive classification algorithms capable of identifying new categories of defects and ensuring accurate classification of labelled categories. Specifically, we developed a multitasking network that utilises Swin-Transformer as the underlying architecture to classify labelled categories and identify novel defective categories. And neuronal activation coverage is designed to detect unseen types of PCBA defects. Furthermore, we propose an end-to-end unsupervised hashing algorithm that incorporates novel category discovery for images classified as previously unseen categories. Finally, we conducted experiments on a variety of different backbone networks in real PCBA defect datasets to demonstrate the effectiveness of our proposed method. Inderscience Publishers - linking academia, business and industry through research

Title: PCBA defect classification in real world with out-of-distribution detection

Authors: Wei Wang; Zhenyi Xu; Yu Kang; Lijun Zhao

Addresses: AHU-IAI AI Joint Laboratory, Anhui University, Hefei, 230601, China ' Key Laboratory of System Control and Information Processing, Ministry of Education, Shanghai, 200240, China; Institute of Artificial Intelligence, Hefei Comprehensive National Science Center, Hefei, 230088, China ' Department of Automation, University of Science and Technology of China, Hefei, 230026, China ' Yangtze River Delta Hart Robot Industry Technology Research Institute, Wuhu, 241060, China

Abstract: Inadequate chip solder joints can significantly impact the overall quality of finalised Printed Circuit Board (PCB) products. Detecting all solder joint defects in real time and adaptively during the actual production process poses a formidable challenge due to the diverse nature of these defects and the limited availability of anomaly data. In contrast to the conventional Printed Circuit Board Assembly (PCBA) defect classification task that is limited to pre-labelled defects, we propose adaptive classification algorithms capable of identifying new categories of defects and ensuring accurate classification of labelled categories. Specifically, we developed a multitasking network that utilises Swin-Transformer as the underlying architecture to classify labelled categories and identify novel defective categories. And neuronal activation coverage is designed to detect unseen types of PCBA defects. Furthermore, we propose an end-to-end unsupervised hashing algorithm that incorporates novel category discovery for images classified as previously unseen categories. Finally, we conducted experiments on a variety of different backbone networks in real PCBA defect datasets to demonstrate the effectiveness of our proposed method.

Keywords: solder joints; defect classification; OOD; out-of-distribution; novel category discovery.

DOI: 10.1504/IJSCIP.2025.146938

International Journal of System Control and Information Processing, 2025 Vol.4 No.3, pp.187 - 203

Received: 28 Sep 2023
Accepted: 26 Mar 2024
Published online: 27 Jun 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article

Title: PCBA defect classification in real world with out-of-distribution detection

Keep up-to-date