Speckle characteristic based auto-focusing approach for diffuse object in digital holography

Abstract

Multi-wavelength digital holography extends the axial measurement range to the order of several tens of microns or more and makes the inspection of optical-rough objects and objects with steep slopes possible. As the optical-rough surface would superimpose speckle field on hologram, the existing focus evaluation criteria frequently fail in indicating correct reconstruction distance. This paper proposes a focus evaluation algorithm for focus detection of diffused surfaces. The average size of the speckle is used as the index and the average size of speckle is evaluated through morphological image processing. Two roughness specimens and a gear tooth are used as test pieces to validate the feasibility of the proposed method. Compared with the existing image focus evaluation method, the proposed morphological average size of speckle approach shows good robustness and sensitivity to various kinds of diffusing surface of mechanical parts.