PCB Fault Detection AI-Based Inspection of Printed Circuit Boards

Abstract

Printed Circuit Boards (PCBs) are essential components of electronic devices, and even minor defects can affect their performance and reliability. Traditional manual inspection is time-consuming and may fail to identify small defects such as missing components, broken copper tracks, soldering errors, and incorrect component placement. To overcome these limitations, the proposed AI-Based PCB Fault Detection System uses computer vision and image processing techniques to analyze PCB images and identify potentially defective regions. The system is developed using Python and OpenCV for image preprocessing and comparison, with an optional machine learning or Convolutional Neural Network (CNN) model for defect detection and classification. The detected results are displayed as a pass/fail status or a possible defect indication, with suspicious regions highlighted wherever supported. This system aims to improve inspection speed, consistency, and quality control while reducing manual effort and rework. However, its reliability depends on image quality, dataset quality, and proper validation.

Introduction

The proposed AI-Based PCB Fault Detection System aims to automate the visual inspection of Printed Circuit Boards (PCBs) using computer vision and artificial intelligence. PCBs are essential components of electronic devices, and even minor defects, such as broken copper tracks, missing components, soldering errors, or incorrect component placement, can affect their performance and reliability.

Traditional PCB inspection relies heavily on manual examination, which is time-consuming, labour-intensive, and prone to human error. To overcome these limitations, the proposed system uses Python, OpenCV, and optional Machine Learning or Convolutional Neural Network (CNN) models to identify potential defects automatically.

1. Aim and Objectives

The main aim is to develop an automated system that analyzes PCB images and detects possible visual defects to improve inspection speed, accuracy, and consistency.

Key objectives include:

• Capturing PCB images using a camera or existing datasets.

• Preprocessing images through resizing, alignment, noise reduction, and enhancement.

• Detecting defects such as missing components, broken tracks, soldering faults, and placement errors.

• Classifying defects and highlighting suspicious regions.

• Displaying inspection results through a simple user interface.

• Evaluating performance using accuracy, precision, recall, and processing time.

2. Literature Survey

Existing PCB inspection techniques include:

• Traditional image processing: Compares PCB images to identify differences and visible defects.

• Machine vision: Automates inspection through image acquisition and feature extraction.

• Machine learning: Classifies defective and non-defective boards using extracted features.

• Deep learning (CNN): Automatically learns defect patterns and improves recognition.

• Object detection: Models such as YOLO, Faster R-CNN, and SSD detect and locate defects.

However, existing approaches may require extensive training datasets, specialized hardware, and significant computational resources. The proposed system focuses on providing a simple and accessible alternative using Python and OpenCV.

3. Proposed Methodology

The system follows seven main stages:

1. Image Acquisition: Collect PCB images using a camera or an existing dataset.

2. Image Preprocessing: Improve image quality through resizing, noise removal, brightness adjustment, and alignment.

3. Feature Extraction: Examine copper tracks, solder joints, component presence, and placement.

4. Defect Detection: Identify suspicious regions using image comparison or trained AI models.

5. Defect Classification: Determine whether the PCB is defective and identify the possible defect category.

6. Result Visualization: Display the inspection status and highlight detected defects.

7. Performance Evaluation: Test the system using accuracy, precision, recall, and processing time.

4. Working of the System

The system operates through the following workflow:

PCB Image Input

Camera or uploaded image

Image Preprocessing

Resizing, alignment and noise reduction

Image Analysis

OpenCV and AI-based detection

Defect Detection and Classification

Identify possible PCB faults

Result Visualization

Highlight suspicious areas

Final Inspection Result

Possible Defect Found / No Visible Defect Detected

5. Expected Results

The proposed system is expected to:

• Automatically identify common visual PCB defects.

• Reduce manual inspection effort and time.

• Improve consistency in quality-control processes.

• Highlight suspicious areas for easier identification.

• Support technicians in preliminary PCB inspection.

Actual performance will depend on image quality, dataset availability, and the selected detection algorithm.

6. Applications

The system can be used in:

• Electronics manufacturing industries.

• PCB quality-control laboratories.

• Electronic repair and maintenance centres.

• Academic and research institutions.

7. Future Scope

The system can be improved by integrating real-time camera inspection, advanced object detection algorithms such as YOLO, additional defect categories, and larger PCB datasets. Further improvements can also focus on increasing detection accuracy and processing speed.

Conclusion

The AI-Based PCB Fault Detection System aims to improve the efficiency and consistency of PCB quality inspection by combining Python, OpenCV, and optional machine learning or CNN models. By automatically analyzing PCB images and identifying suspicious regions, the system can assist technicians in detecting visible defects at an early stage, potentially reducing rework and improving manufacturing quality. Its performance depends on image quality, dataset diversity, and the effectiveness of the selected detection algorithm. In the future, the system can be enhanced through real-time camera integration, improved defect classification, and more accurate localization techniques. Although it supports visual inspection, additional electrical testing and expert verification are necessary to identify faults that cannot be detected from images alone.

References

[1] badboysm890. Pcb defect detection using yolov5. GitHub repository, 2023. [Online]. Available: https://github.com/badboysm890/PCB-Defect-Detection. [2] JiaLim98. Yolo-pcb defect dataset & detection. GitHub repository, 2023. [Online]. Available: https://github.com/JiaLim98/YOLO-PCB. [3] PCBWay. Pcb flying probe test: Principles, advantages, and applications. PCBWay Blog, 2023. [Online]. Available: https://www.pcbway.com/blog/PCBManufacturingInformation/PCBFlyingProbeTestPrinciplesAdvantagesandApplicationsbc222b0f.html. [4] ProtoExpress. How flying probe testing works for pcb assembly. ProtoExpress Blog, 2023. [Online]. Available: https://www.protoexpress.com/blog/how-flying-probe-testing-worksfor-pcb-assembly/. [5] ResearchGate. A review: Comparison of performance metrics of pretrained models for object detection using the tensorflow framework - scientific figure. ResearchGate, 2025. [Online]. Available: https://www.researchgate.net/figure/Comparison-of-framesprocessed-per-second-FPS-implementing-the-Faster-R-CNN-R-FCN-SSDfig6342570032 [Accessed : 4Nov2025]. [6] ryanq04. Automated pcb tester with stm32 and python gui. GitHub repository, 2022. [Online]. Available: https://github.com/ryanq04/automatedpcbtester. [7] H. Wang et al. Lightweight improved yolov5 algorithm for pcb defect detection. The Journal of Supercomputing, 2024. [Online]. Available: https://link.springer.com/article/10.1007/s11227-024-06739-w. [8] S. Zhang et al. Yolo-pdd: A novel multi-scale pcb defect detection method using deep representations with sequential images. arXiv preprint, 2024. [Online]. Available: https://arxiv.org/abs/2407.15427. [9] K. Eves and J. Valasek, Adaptive control for singularly perturbed systems examples, Code Ocean, Aug. 2023. [Online]. Available: https://codeocean.com/capsule/4989235/tree [10] S. Liu, Wi-Fi Energy Detection Testbed (12MTC), 2023, gitHub repository. [Online]. Available: https://github.com/liustone99/Wi-Fi-Energy-Detection-Testbed-12MTC

Copyright

Copyright © 2026 Yogini Diwate, Tejas Chamat, Bhavesh Mesharam, Darshak Kakde, Mohd Faik. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.