AI & Computer Vision
AI Camera Inspection QC Check
Computer-vision inspection workflow for defect visibility and operator-facing quality-control support.
Workflow Illustration
This inspection workflow uses camera capture and AI inference to support quality control in a more consistent way than manual visual checking alone. Image capture, defect localization, OK-NG decision, and evidence storage help reduce subjective review and provide clearer operator-facing inspection results.
Business Problem
Manual visual inspection is repetitive and can be inconsistent when operators need fast defect detection in real operational conditions.
My Contribution
Built and presented a QC workflow using Python, YOLO, and OpenCV, with image and video-based evidence to show inspection support scenarios and defect visibility.
Operational Flow
- Parts enter a controlled inspection zone with camera and lighting support.
- Captured images are processed by the computer-vision model for defect detection.
- Detection results are shown as operator-facing OK-NG guidance with visual evidence.
- Inspection snapshots support traceable review beyond manual notes alone.
Key Capabilities
- Camera-based quality inspection support.
- YOLO-driven defect detection and visualization.
- Operator-facing OK-NG inspection visibility.
- Image-based evidence for review and traceability.
Tech Stack
Python, YOLO8, OpenCV.
Current Status
Status: Computer-vision workflow demonstration focused on defect visibility, operator support, and image-based inspection evidence.
Includes sanitized media assets for project illustration. The source repository remains private.