Digital holographic particle validation via complex wave

Abstract

The inability to distinguish between particle images and noise in holographic reconstruction of dense particle fields hampers the advancement of holographic particle diagnostic techniques including holographic particle image velocimetry. We developed a method to separate particles from the noise by unlocking a unique particle signature in the complex reconstructed field. This complex-wave signature is present in digital particle holograms recorded at any scattering angle. Simulations of single and multiple particle holograms, as well as preliminary laboratory particle-field experiments, not only demonstrated the existence of the particle signature but also evaluated its ability to remove noise. Regardless of particle seeding density, scattering angle of hologram recording and particle size range, the particle identification/validation routine consistently provides >50% removal of "bad" particles and <8% of good particles.

Computational imaging and automated identification for aqueous environments

“…Note that a darker cone extends out beyond the point object, so that points near to the object have minimum intensities at defocus greater than zero, and thresholding is necessary to limit the focus metric to the correct region. [89], [274]. Each of these methods is developed for use with particles.…”

“…(Note that particle metrics such as de Jong's complex signature [89] or Pan's minimum imaginary component variance [274] are inappropriate for these cases since the object cannot be considered to be point-like.) The assumption of minimum intensity is valid when the particles go out of focus quickly and the noise level is low, so that noise does not signi…cantly impact the intensity and the object edges have a low value for only a few reconstruction slices.…”

“…Holographic particle imaging velocity tracks particles through 3D volumes to produce estimates of the ‡ow …eld [160], [273], [290], [240], [328], [441], with a number of speci…c algorithms developed for dealing with particle localization [274], [89], [435], [83], [325], [348], [11], [13], [336] and pairing between images [327], [432]. Early …lm holography sought to measure particle size distributions in the environment [377], a trend continued with digital holography [272], [321], [253] (and Section 4.3).…”

“…The real and imaginary components are useful as diagnostics, as components of focus metrics [89], [274], as representations of the objects, or for computing the quantitative phase of imaged objects [75]. The squared magnitude gives the intensity which would be observed using optical reconstruction methods.…”

Axial particle positioning by wavefront parameterization using Chebyshev polynomials and off-axis digital holography

“…A metric is therefore most commonly applied to determine the axial position. Reported metrics have been based on intensity [2], intensity gradients [3] or a combination of both [4], complex metrics have also been reported [5,6]. We have in a previous paper showed that the axial particle position can be found by tracking the wavefront curvature [7].…”