Automated Visual Inspection for Precise Defect Detection and Classification in CBN Inserts

Area of Science:

• Materials Science and Engineering
• Manufacturing Technology
• Computer Vision and Image Processing

Background:

• Cubic Boron Nitride (CBN) inserts are critical in precision manufacturing due to their exceptional hardness.
• Surface defects on CBN inserts can significantly degrade product integrity and performance.
• Existing defect detection methods may lack the speed and accuracy required for automated production lines.

Purpose of the Study:

• To develop and validate an automated machine vision system for detecting and classifying surface defects on CBN inserts.
• To evaluate the performance of various defect detection algorithms for CBN insert inspection.
• To create a robust and efficient system for enhancing quality control in high-speed manufacturing.

Main Methods:

• Integration of an optical bracket, high-resolution industrial camera, precise lighting, and an advanced development board for image acquisition.
• Application of digital image processing techniques for defect identification and categorization.
• Comparative analysis of multiple defect detection algorithms, considering parameter tuning and dataset diversity.

Main Results:

• The developed system achieves a detection accuracy exceeding 90% for multiple defect types.
• The system demonstrates a tooth surface recognition efficiency of three frames per second.
• The front and side cutting surfaces of the tool are effectively captured and analyzed within each frame.

Conclusions:

• The proposed machine vision system offers a scalable and reliable solution for automated surface defect detection on CBN inserts.
• This technology significantly improves quality control in automated, high-speed precision manufacturing environments.
• The system's high accuracy and efficiency pave the way for enhanced production line monitoring and defect management.