Structure-Guided Mixed Masked Pretraining and Spatial Continuity Regularization for Printed Circuit Board Defect Detection
2026-06-02 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors developed a two-step method to find tiny defects on printed circuit boards (PCBs), which are hard to spot because they're small and blend into complex backgrounds. First, they trained a computer model to understand PCB structures by hiding parts of images and teaching the model to fill in the gaps. Then, they fine-tuned the model to better locate defects by encouraging it to group related defect spots together. Their experiments showed this approach works better than other strong methods in finding PCB defects reliably.
Printed Circuit Board (PCB)Defect DetectionAutomated Optical Inspection (AOI)Masked PretrainingSparse ConvolutionSpatial Continuity RegularizationFine-tuningMean Average Precision (mAP)Structure-guided Masking
Authors
Peitong Wang, Nuo Wang, Enxin Qin, Chengjin Yu, Hanyu Xuan, Yuanting Yan
Abstract
Printed circuit board (PCB) defect detection is an essential part of automated optical inspection (AOI); yet it remains challenging in practice because many defects are tiny, low-contrast, and embedded in dense circuit backgrounds. To address these issues, this paper presents a two-phase PCB defect detection framework that combines structure-guided mixed masked pretraining with spatial continuity regularization. In the pretraining stage, we design a sparse convolutional masked pretraining scheme to exploit unlabeled PCB images, where structure-guided mixed masking is used to construct informative masked inputs. The sparse convolutional reconstruction pipeline suppresses invalid responses from masked regions and enables the detector backbone to infer missing PCB structures from visible conductive patterns, thereby learning PCB structural priors. In the fine-tuning stage, the pretrained backbone is transferred to the downstream defect detection task. For the task, a spatial continuity regularization term is introduced during fine-tuning. This term constrains dispersed positive predictions assigned to the same defect instance and promotes more compact localization on elongated defect regions. Experiments on the DsPCBSD+ dataset show that the proposed method achieves 85.5% mAP0.5 and 52.3% mAP0.5:0.95, outperforming several strong baseline detectors. Ablation studies and qualitative results further confirm the effectiveness of the proposed framework for robust PCB defect detection in industrial AOI scenarios.