BoardVision: Deployment-ready and Robust Motherboard Defect Detection with YOLO+Faster-RCNN Ensemble

TL;DR

BoardVision tackles assembly-level motherboard defect detection by benchmarking YOLOv7 and Faster R-CNN on the MiracleFactory dataset and introducing a lightweight Confidence–Temporal Voting (CTV) ensemble to balance precision and recall. The method achieves higher F1 and stable precision than either model alone, and demonstrates robustness under lighting, sharpness, and orientation perturbations, all while providing a deployable GUI for operator use. The work delivers a reproducible framework that bridges research benchmarks with factory-floor QA, and highlights the importance of robust, interpretable ensemble strategies for deployment. Overall, BoardVision advances practical, deployment-ready vision systems for electronics manufacturing, combining rigorous benchmarking, robustness analysis, and an operator-centric toolchain.

Abstract

Motherboard defect detection is critical for ensuring reliability in high-volume electronics manufacturing. While prior research in PCB inspection has largely targeted bare-board or trace-level defects, assembly-level inspection of full motherboards inspection remains underexplored. In this work, we present BoardVision, a reproducible framework for detecting assembly-level defects such as missing screws, loose fan wiring, and surface scratches. We benchmark two representative detectors - YOLOv7 and Faster R-CNN, under controlled conditions on the MiracleFactory motherboard dataset, providing the first systematic comparison in this domain. To mitigate the limitations of single models, where YOLO excels in precision but underperforms in recall and Faster R-CNN shows the reverse, we propose a lightweight ensemble, Confidence-Temporal Voting (CTV Voter), that balances precision and recall through interpretable rules. We further evaluate robustness under realistic perturbations including sharpness, brightness, and orientation changes, highlighting stability challenges often overlooked in motherboard defect detection. Finally, we release a deployable GUI-driven inspection tool that bridges research evaluation with operator usability. Together, these contributions demonstrate how computer vision techniques can transition from benchmark results to practical quality assurance for assembly-level motherboard manufacturing.

Figures (10)

• Figure 1: BoardVision pipeline: YOLOv7 and Faster R-CNN outputs are reconciled by the Confidence–Temporal Voting ensemble and visualized through a GUI for operator-facing deployment.
• Figure 2: Ground truth annotations for the motherboard test image used in Figures \ref{['fig:false_case_examples']} and \ref{['fig:voter_qual_1']}.
• Figure 3: Annotated error cases for YOLOv7 and FRCNN. False positives (FP) and false negatives (FN) are annotated in red to illustrate specific failure modes.
• Figure 4: Voter Predictions for motherboard test image.
• Figure 5: Live inference on a low-light, low-resolution stream. YOLOv7 misses a CPU fan, Faster R-CNN detects it with high confidence ($p=0.98$), and the voter ensemble fuses the result into the final decision (yellow box).