PCB Inspection System Using Computer Vision

The PCB Visual Inspection Studio is a Windows-based Automated Optical Inspection (AOI) project that compares a test PCB against a known-good golden reference board and identifies visible differences using classical computer vision techniques.

The system performs image registration, difference analysis, morphological processing, connected-component analysis, and heuristic defect classification to detect potential PCB abnormalities.

Key Features

• Golden-board vs test-board comparison
• Missing component detection
• Misplaced component detection
• Solder-region difference detection
• Scratch detection
• Other visible defect detection
• ORB / AKAZE image registration
• RANSAC homography alignment
• Difference heatmap and defect overlay
• Ignore ROI support
• Configurable inspection recipes
• Webcam / USB camera capture
• Inspection history
• PASS/FAIL analysis
• CSV, JSON, PNG and HTML report export
• Built-in synthetic demo boards
• Completely offline operation

The inspection pipeline aligns the test image with the reference board, generates a difference map, applies thresholding and morphology, identifies connected components, and assigns heuristic defect labels.

Technologies

Python | PyQt | OpenCV | NumPy | Computer Vision

What You Get

• Complete project source code
• PCB inspection application
• Computer-vision processing modules
• Defect classification logic
• Camera capture functionality
• Recipe and configuration system
• Inspection history
• Report export functionality
• Demo/sample boards
• Test suite

Ideal For

B.Tech/M.Tech Projects • Electronics & PCB Projects • Computer Vision Projects • AI/ML Projects • Industrial Automation Projects • Engineering Research

Important Note

This is a classical computer-vision AOI project, not a deep-learning detector. Inspection accuracy depends significantly on consistent lighting, camera positioning, board alignment, and image quality.