Three dimensional (3D) imaging of the tympanic membrane (TM) has been carried out using a traditional otoscope equipped with a high-definition webcam, a portable projector and a telecentric optical system. The device allows us to project fringe patterns on the TM and the magnified image is processed using phase shifting algorithms to arrive at a 3D description of the TM. Obtaining a 3D image of the TM can aid in the diagnosis of ear infections such as otitis media with effusion, which is essentially fluid build-up in the middle ear. The high resolution of this device makes it possible examine a computer generated 3D profile for abnormalities in the shape of the eardrum. This adds an additional dimension to the image that can be obtained from a traditional otoscope by allowing visualization of the TM from different perspectives. In this paper, we present the design and construction of this device and details of the imaging processing for recovering the 3D profile of the subject under test. The design of the otoscope is similar to that of the traditional device making it ergonomically compatible and easy to adopt in clinical practice.
The most known and used phase shifting interferometry (PSI) demodulation methods are one-dimensional temporal linear systems. These methods use the information of the interferogram sequence at a single pixel to recover the modulating phase. Accordingly, scanning all pixels, we obtain the two-dimensional (2-D) modulated phase sought. As PSI demodulation methods do not take into account spatial information, these methods cannot remove unwanted harmonics or noise from the interferogram image space (spatial domain). To remove these unwanted artifacts from the image space, spatial information must be included in the demodulation model. We are going to show that the well-known least-squares system for PSI can be used as a full-field 2-D linear system that uses the temporal and spatial information in conjunction in order to recover the modulating phase while removing noise, unwanted harmonics, and interpolating small empty sections of the image space all in the same process with a low computational time.
Recent advances in the acquisition of <i>in-vivo</i> high resolution retinal images through the use of Adaptive Optics (AO) have allowed the identification of cellular structures such as cones and rods, in and out of the fovea, in such a way that their histological characteristics can be studied <i>in-vivo</i> and later compared to data obtained <i>post-mortem</i>. In this work, an algorithm is proposed for the detection of photoreceptors; it consists of two stages: Early Cell Detection (ECD), to detect all candidate cells, and Refinement of Cell Detection (RCD), to reduce over-detection of photoreceptors. The algorithm has been tested using synthetic and real images, the latter acquired with an Adaptive Optics Scanning Light Ophthalmoscope (AOSLO). The proposed algorithm was compared against the one developed by Li and Roorda, and both algorithms were tested on synthetic and real images, yielding similar algorithm performance on both kinds of images when they had only cones; however, the algorithm developed by Li and Roorda, when applied to real images having cones and rods, identifies photoreceptors in vascular tissue, in addition to showing low rod detection.
In this work, we develop a regularization technique to demodulate a phase-shifting interferogram sequence with arbitrary inter-frame phase shifts. With this method, we can recover the modulating phase and inter-frame phase shifts in the same process. As all phase-shifting algorithms, the assumption is that the wavefront under test does not change over time, but the introduction of phase-shifts can vary in a nonconstant way. A notable characteristic of this demodulation method is that not only can it recover the modulating phase, but it is also capable of filtering-out large amounts of corrupting noise. We will show numerical experimental results and comparisons with another already published method to see the performance of the demodulation technique developed herein.
Pixelated phase-mask (PPM) interferometers have become an industry standard for instantaneous
phase-shifting interferometry. In commercially available PPM interferometers, an array with 2x2
unit-cells is used, which codify up-to 4 phase-steps within a single PPM interferogram. Recently we
have shown that such 2x2 unit-cell arrays allows a harmonic rejection as good as the 4-step leastsquares
phase-shifting algorithm (LS-PSA); this harmonics rejection is relatively-low and may not
be enough to correctly demodulate some severely intensity distorted fringe patterns. In previous
works we have proposed a new PPM with a 3x3 unit-cell to improve the harmonics rejection of the
2x2 array. With this new 3x3 unit-cell one is able to reject as many harmonics as with a 9-step LS-PSA<sup>10</sup>.
In this paper we are extending the analysis of MxN unit-cell synchronous demodulation of
PPM. The new results allow us to answer some important open questions about the method: for a
given configuration, which harmonics cannot be rejected and why? Why, prior to low-pass filtering,
we observe multiple copies of the interferogram’s spectrum and what does this imply? We believe
these preliminary results are important contributions towards a formulation of a general theory MxN
unit-cell pixelated carrier interferometry.
Here, we are going to show a 2D recursive phase unwrapping system that unwraps the phase and cleans noise
in the same process. This recursive phase unwrapping system behaves as an Infinite Impulse Response (IIR)
system and we will show its stability and frequency analysis in 1D. The unwrapping results shown here, are
compared with the well known least squares phase unwrapping method. The recursive phase unwrapping system
shown here is robust to noise and it can be used to unwrap modulus 2π phase maps following simple row by row
The present work shows preliminary results of a phase unwrapping technique used in interferometry. Wrapped
phase maps are the result of the modulus 2π ambiguities caused for the phase recovery function arctan. Here
we present a recursive n-order phase unwrapping system that removes the ambiguities, it's robust to noise and
fast. The system is able to recover the unwrapping phase in presence of high noise, according to stability of the
system that can be controlled. This high noise causes line sequential integrations of phase differences to fail.
The system is not numerically-heavy in comparison with other methods that tolerate the noise. The application
areas of the system can be: optical metrology, magnetic resonance, and those imaging systems where information
is obtained as a demodulated wrapped phase map.
Phase-shifting algorithms are methods used for recovering the modulating phase of an interferogram sequence
obtained by Phase Stepping Interferometry (PSI) techniques. Typically, the number of interferograms in a PSI
sequence is from 3 to around 9 interferograms, although we can find algorithms that works with more than
9 interferograms. In this paper, we are going to show the analysis and design of phase-shifting algorithms
from the point of view of the linear systems paradigm from digital signal processing. We will show how this
paradigm describes in a general fashion the phase-shifting algorithm systems, and how we can easily design
tunable phase-shifting algorithms using this simple scheme.
In our Optical Metrology laboratories, we deal with the problem of demodulating temporal sequences
of interferograms. These sequences of interferograms are obtained by means of optical
testing of transient events using an Electronic Speckle Pattern Interferometry system (ESPI). It is
well known that using Phase Stepping Interferometry techniques (PSI), one can obtain the modulated
interferogram phase with at least three equally temporal phase shifted interferograms. To
obtain these three (or more than three) phase shifted interferograms with a conventional ESPI array,
it is necessary to have a static object under test. On the other hand, if the object under test is not
static, we can make the analysis using dual-pulse subtraction ESPI, introducing a spatial frequency
carrier. However, in our case, we will use a conventional ESPI technique, with a continuous laser
and without a frequency carrier. Thus, as we pretend to analyze transient deformations or events
without frequency carrier, we can not use the demodulation methods used in dual-pulse subtraction
ESPI, nor PSI techniques because it results almost impossible to take the least amount of interferograms
with the required linear phase shift (or temporal carrier) among them. To accomplish
this, it will be necessary look for alternatives to demodulate temporal sequences of interferograms
without a frequency carrier and without linear phase-shifting. Here, we present the groundwork
aimed at demodulating sequences of interferograms without a frequency carrier, where traditional
PSI techniques are unable to detect the phase correctly.
We propose a new approach to demodulate a single fringe pattern with closed fringes by using Local Adaptable Quadrature Filters (LAQF). Quadrature filters have been widely used to demodulate complete image interferograms with carrier frequency, however these have never being used to demodulate complete image interferogramas with out carrirer (with closed fringes). The idea, in this paper, is to demodulate the fringe pattern sequentially, using a fringe following scanning strategy. In particular we use linear <i>robust quadrature filters</i> to obtain a fast and robust demodulation method for single fringe pattern images with closed fringes. The proposed LAQF method does not require a previous fringe pattern normalization. Some tests with experimental interferograms are shown to see the performance of the method along with comparisons to its closest competitor, which is the Regularized Phase Tracker (RPT), and we will see that this method is tolerant to higher levels of noise.
In this work we present a sequential technique for temporal fringe pattern demodulation without
carrier. The technique presented here, uses a temporal frequency estimator to obtain the temporal
phase from a interferogram sequence. The restriction used to estimate the frequency is based on
second order potentials in order to obtain the temporal phase as a phase function in an space C<sup>2</sup>.
The importance of this technique to demodulate temporal fringe patterns without carrier is mainly
the simplification of experimental optical arrays in laboratory, where the experimental nature make
it difficult to introduce a carrier frequency. This work present an on development technique for
temporal demodulation, however, its projection on future work give us the possibility to obtain a
robust demodulation method for temporal interferograms. To demonstrate the performance of this
temporal fringe pattern demodulation technique, we are going to show a simulated interferogram
sequence to demodulate it with this technique.
Sequential methods like the regularized phase tracker (RPT) are commonly used for fringe pattern
demodulation with closed fringes. The only drawback of the RPT method is the necessity to implement
a two-dimensional (2D) fringe following scanning in order to obtain the expected modulated
phase. In this article we present a new method to demodulate single fringe patterns with closed
fringes which use a simple 2D row by row scanning strategy. This is an important contribution
because the 2D row by row scanning is extremely fast and easy to implement unlike the fringe following
scanning. We have called this method the <i>phase curvature tracker </i>(PCT) because it uses the
frequency curvature as regularizer to obtain the expected phase as a C<sup>2</sup> function with continuous