A Deep Learning Driven Approach For Automated Identification And Classification Of Defects In Printed Circuit Boards

• Author(s):

  Roshini H | Thigazhvazhaki B | Dr. S. Sathiya Priya

• Keywords:

  PCB Defect Detection, YOLOv8, Deep Learning, Real-Time Inspection, Computer Vision, Flask Web Application, Automated Optical Inspection (AOI), Surface Defect Classification, Industrial Automation, Smart Manufacturing.

• Abstract:

  In The Fast-paced Electronics Manufacturing Industry, Ensuring The Quality Of Printed Circuit Boards (PCBs) Is Essential For Producing Reliable And High-performance Devices. Traditional Methods For PCB Defect Detection, Such As Manual Visual Inspection And Rule-based Computer Vision, Are Often Limited By Their Inability To Detect Small, Complex, Or Subtle Defects, Leading To High False-positive Rates And Inefficiencies. This Paper Introduces A Deep Learning-driven Approach For The Automated Identification And Classification Of Defects In PCBs, Leveraging The Power Of YOLOv8, A State-of-the-art Object Detection Model. The Proposed System Is Capable Of Accurately Detecting A Wide Range Of Defects, Including Missing Holes, Mouse Bites, Open Circuits, Shorts, Spurious Copper, And Spurs, With Real-time Performance And High Precision. The Model Is Trained On A Diverse Dataset Of PCB Images, And The Detection System Is Integrated Into An Intuitive Web Application Built With Flask, Allowing Users To Easily Upload PCB Images For Instant Analysis. Experimental Results Show That The System Achieves A Mean Average Precision (mAP) Greater Than 90%, Significantly Outperforming Traditional Approaches In Both Accuracy And Inference Speed. This Approach Not Only Automates The Defect Detection Process But Also Provides A Scalable And Efficient Solution For Quality Control In PCB Manufacturing. Future Improvements To The System Will Focus On Multi-layer PCB Inspection, Incorporating Advanced Imaging Techniques Like X-ray And Infrared Imaging, As Well As The Development Of AI-driven Predictive Maintenance Features And Edge Computing For Real-time, On-device Inference.

---

Other Details

• Paper id:

  IJSARTV11I5103583

• Published in:

  Volume: 11 Issue: 5 May 2025

• Publication Date:

  2025-05-16

Download Article