A reflection heat source including a radiator as well as an aluminum plate is designed, and the temperature
field of the aluminum plate is used as the tested object. The reflection lensless Fourier transform (LFT) digital
holography is performed to measure the temperature field distribution. For the comparison, the temperature
measurement system within the radiator is used to measure the temperature distributions. The results obtained
by these two methods are in good agreement, which demonstrates that the digital holography method is valid
for the measurement of the temperature distribution.
There exist some ambiguities about the depth of view (DOV) and depth of focus (DOF) of digital holographic imaging
system (DHIS). Based on the principle of digital holography, the DOV and the DOF of the DHIS is analyzed in detail for
the first time to the best of our knowledge. For the four common configurations for recording digital holograms, the axial
complex amplitude and the intensity distributions of the reconstructed optical field of a point object are deduced
respectively using Fresnel diffraction formula. According to the same criterion in common coherent imaging system of
the lens (CCISL), the DOF of the DHIS is obtained. The results show that the DOV and the DOF of the DHIS are related
not only to the optical wavelength and the numerical aperture of the CCD camera but also to the offset of the reference
light wave. The DOF of the CCISL is found larger than that of the digital in-line holographic system. But the DOF of the
digital off-axis lensless Fourier transform holographic imaging system, which is depended strongly on the offset of the
reference point source, may be larger or smaller than that of the CCISL. The computer simulation results confirm the
validity of the theoretical analysis.
A simple holographic high-resolution imaging system without pre-magnification, which is based on off-axis lensless Fourier transform configuration, has been developed. Experimental investigations are performed on USAF resolution test target. The method based on angular spectrum theory for reconstructing lensless Fourier hologram is given. The reconstructed results of the same hologram at different reconstructing distances are presented for what is to our knowledge the first time. Approximate diffraction limited lateral resolution is achieved. The results show that the angular spectrum method has several advantages over more commonly used Fresnel transform method. Lossless reconstruction can be achieved for any numerical aperture holograms as long as the wave field is calculated at a special reconstructing distance, which is determined by the light wavelength and the chip size and the pixel size of the CCD sensor. This is very important for reconstructing an extremely large numerical aperture hologram. Frequency-domain spectrum filtering can be applied conveniently to remove the disturbance of zero-order. The reconstructed image wave field is accurate so long as the sampling theorem is not violated. The experimental results also demonstrate that for a high quality hologram, special image processing is unnecessary to obtain a high quality image.
Digital holography is particularly well suited for characterization of microstructure such as surface shape, surface nanostructures and surface roughness. The direct availability of both amplitude and phase information offers a range of versatile processing techniques that can be applied to complex field data, including phase imaging, which is particularly straightforward in digital holography. Based on digital off-axis lensless Fourier transform holographic configuration, the principle of topography by digital holography is analyzed. The wavefront deformation created by numerical Fresnel reconstruction is then studied. Thus the model of phase mask which is a parabolic function and related to the recording parameters is built. The phase mask can be obtained automatically by using digital hologram itself to evaluate automatically and accurately the values of the parameters involved by a curve-fitting procedures applied to the phase data extracted along two profiles, which are defined in the region where sample contributions are constant. The reconstructed wave field can be simply obtained by multiplying the phase mask with Fourier transform of the hologram. Both theoretical analysis and the simulation results show that the procedure proposed by Colomb is more suitable for phase reconstruction of digital off-axis lensless Fourier transform holography. Moreover, the correcting procedure can be applied to compensate high-order aberrations.
Based on the point spread function of the off-axis Fresnel digital holographic system with pre-magnification, the phase
aberration introduced by microscope objective is obtained. Using a collimated light as reconstructing wave, the phase
aberration introduced by the difference between the reconstructing wave and the reference recording wave is analyzed.
This method is very simple, and it is very different from the one proposed by T. Colomb <i>et al</i>. The phase mask that can
be used to compensate automatically the phase aberration in the phase reconstruction is obtained. According to the
principle of digital holography, it must be pointed out that for Fresnel digital holography, the recording distance must be
determined accurately before compensating automatically the phase aberration. This is done by an automatic focusing
procedure, which is based on image-gray-entropy-method. The simulation result for a special three-dimensional micro
object, which is polluted by a random noise, is presented. The percentage error of the reconstructing distance obtained by
the focusing procedure is below 0.6 for SNR = 25. Then an automatic aberration compensation procedure, which is the
same as that one proposed by T. Colomb <i>et al</i>., is applied to reconstruct the phase image. The results show that for a
weak noise the above method is very effective; for a stronger noise the procedure described here is applied iteratively,
starting from the initial values provided by the first evaluation; while for a very strong noise the procedure fails at all.
Moreover, after applying a median filter to the primary reconstructed phase image, the aberration of the phase image
obtained by further iteration decreases, at the same time the noise is strengthen.
Digital holography is a whole-field, non-contact, and highly sensitive interferometric imaging and testing technology. It
is more suitable for microscopic measurement owing to digitalization and flexibility in holograms recording, storage,
reconstruction and transmittance. This paper analyzes the factors which lead to the phase aberrations in the off-axis
lensless Fourier transform digital holography firstly. Then a method, which is obtained by borrowing ideas from T.
Colomb, is presented to correct the phase aberrations automatically during the numerical reconstruction. This is
implemented by multiplying a phase mask directly with the reconstructed wave field. The phase mask is obtained by an
iterative procedure computing automatically without the pre-knowledge of the physical parameters, such as the off-set of
the reference point source and the recording distance. This method enables one to reconstruct the relative correct and
accurate phase-contrast image, even in the presence of the noise, which is needed to be smoothed by a median filter. In
order to achieve an accurate phase image, the procedure described here is applied iteratively, starting from the initial
values provided by the first evaluation. We present and analyze the simulation results of the phase images based on a
special three-dimensional micro object. The results show that for a weak noise the above method is very effective; while
for strong noise the common phase-unwrapping method must be applied. This indicates that it is very important to record
high quality hologram and to suppress the noise in phase data.
We present an analysis of the reconstruction through any intermediate plane by using the cascaded Fresnel diffraction
integrals. The realization of the integral with the single fast-Fourier-transform (FFT) imposes inherently the constraints
on the pixel resolution and the field of view of the computation window in the intermediate plane. On the other hand,
according to the optical diffraction field of the band-limited object, the size of the diffraction field including most
energy exhibits a change with the intermediate plane in a different way. Since the field of view of the computation
window is usually smaller than the corresponding optical diffraction field, then the simply cascaded FFT-Fresnel
diffraction integral algorithms will lose the resolution and degrade the image quality. It is proposed to improve the
reconstruction by compensating the field to be close to the optical diffraction field and using the new field as the
computation window in the intermediate plane. The numerical simulation and the experimental data-based calculation
are performed and demonstrate our proposal.
The pre-magnification digital holography with a microscopic objective (MO) has become an important mean for imaging and measuring of microscopic objects. Based on Fresnel diffraction theory, the point spread function of digital off-axis Fresnel holographic system with pre-magnification is firstly deduced in detail. The lateral resolution of this imaging system is then analyzed. The amplitude distributions of the point spread functions of the MO, the holographic CCD and the whole imaging system are simulated respectively. Then, the matching requirements between MO and the holographic CCD are discussed. Some ambiguities about the resolution limitation in existing literatures are clarified. The results show that only when the image resolution of the holographic CCD is not below the imaging resolution of the MO, the resolution of overall system is dependent on the numerical aperture of the MO. Otherwise, it is dependent on the numerical aperture of the CCD. It is optimum to make the imaging resolution of the MO approach its limitation and be equal to or less than the resolution of the holographic CCD a little. The simulation results verify the validities of the theoretical analysis.