Research on Interference Filter's Use in PCB Machine Vision Inspectio

Research on Interference Filter's Use in PCB Machine Vision Inspectio

Citation

梁杰民 (LIANG Jie-min), 郭宝平 (GUO Bao-ping), 杨 欧 (YANG Ou), & 胡 涛 (HU Tao). (2007年6期). 干涉滤光片在 PCB机器视觉检测中的应用研究 [Research on Interference Filter's Use in PCB Machine Vision Inspection]. 科学技术与工程 [Science Technology and Engineering], 7(6), 1048-1050.

Keywords

  • 干涉滤光片 / Interference Filter
  • PCB / 印制电路板 / Printed Circuit Board
  • 机器视觉检测 / Machine Vision Inspection
  • 检测效率 / Inspection Efficiency
  • 图像清晰度 / Image Clarity / Definition
  • 图像处理 / Image Processing
  • 检测误判 / Detection Misjudgment / False Alarm
  • 漏检率 / Undetected Rate / Miss Rate
  • 灰度差 / Gray Level Difference
  • 光谱响应 / Spectral Response
  • 中心波长 / Center Wavelength
  • 透射光谱 / Transmission Spectrum
  • 光谱宽度 / Spectral Bandwidth / Half-Bandwidth
  • CCD照相机 / CCD Camera
  • 图像采集 / Image Acquisition

Brief

This article proposes using interference filters to enhance the efficiency of PCB machine vision inspection by improving image clarity, simplifying processing, and reducing detection errors through optimized contrast between conductors and the substrate.

Summary

This article proposes using interference filters to enhance PCB machine vision inspection by selectively filtering light wavelengths to increase the contrast (gray level difference) between the conductive traces and the base material. This leads to clearer images, simpler image processing, reduced false alarms, and fewer missed defects in PCB automatic inspection. The study analyzes the characteristics of interference filters and how they can be used to optimize the spectral response for different PCB types. Experiments show that selecting the appropriate center wavelength for the interference filter can significantly improve image quality and detection accuracy. However, the method requires sufficient light intensity and may not be effective for PCBs where the conductor and substrate have similar reflective properties across the visible spectrum.

Origin:

Back to blog