Surface roughness measurement by digital speckle correlation

Ichirou Yamaguchi, Koichi Kobayashi, Leonid Yaroslavsky

Research output: Contribution to journalConference articlepeer-review

Abstract

Noncontacting measurement of roughness of solid surfaces by digital speckle correlation of video signals is reported. Speckle patterns appearing in the diffraction field of a laser-illuminated sample are taken by a CCD before and after the change of the incident angle and their cross-correlation peak is calculated as a function of the change from which surface roughness can be evaluated. The theoretical cross-correlation function is derived that describes speckle displacement and decorrelation due to the changes. In the theory the surfaces are assumed to affect only the phase of the incident light in proportion to surface profiles. The decorrelation curve against speckle displacement that is proportional to the change of the incident angle depends on the root-mean-square surface when it is larger than wavelength and when the correlation length of the roughness is much smaller than the spread of the incident beam. We developed an instrument that provides the decorrelation curve in a few tens of seconds by installing a real time correlation device based on phase-only-correlation algorithm. Various roughness standards of molded metal were measured with both the instrument and a stylus roughness meter. Good agreement has been observed between the results for the surface roughness between a few and a few tens of micrometers.

Original languageEnglish
Article number19
Pages (from-to)178-189
Number of pages12
JournalProceedings of SPIE - The International Society for Optical Engineering
Volume5532
DOIs
StatePublished - 2004
EventInterferometry XII: Applications - Denver, CO, United States
Duration: 4 Aug 20045 Aug 2004

Keywords

  • Correlation analysis
  • Laser speckle
  • Stylus instrument
  • Surface roughness

Fingerprint

Dive into the research topics of 'Surface roughness measurement by digital speckle correlation'. Together they form a unique fingerprint.

Cite this