Journal of The Optical Society of America A-optics Image Science and Vision最新文献

A deep learning-based phase demodulation algorithm is proposed for measuring in-plane displacements in conjugate orbital angular momentum (OAM) interferometry. The phase demodulation hybrid neural network (PDHNN) is designed to directly demodulate petal-shaped interferograms in a single step. PDHNN employs a custom ResNet-transformer architecture with deformable convolutions and attention mechanisms to extract rotation-sensitive features from petal-shaped interferograms for robust phase demodulation. The algorithm has been validated using both simulated and experimental data. Experimental results show that the demodulation accuracy reaches 91.60% within an error margin of 1°, and within a 0.1° error range, the average displacement error is 0.13 nm, demonstrating high robustness and stability in noisy conditions.

We introduce a technique using closure amplitudes, inspired by radio interferometry, to determine, with high angular resolution, the two-dimensional profile of a light beam using an interferogram from a non-redundantly masked aperture. Previous techniques required multiple interferograms or accurate estimates of the non-uniform illuminations across the aperture. In contrast, our method, using closure amplitudes, avoids the need to estimate the aperture illuminations while determining the two-dimensional beam shape from a single interferogram. The invariance of closure amplitudes to even time-varying aperture illuminations makes it suitable for longer averaging intervals, with the potential to reduce data rates and computational overheads. By using data from the ALBA synchrotron light source to validate the method and its results against existing methods, this paper represents, to our knowledge, the first real-world application of closure amplitudes to directly determine the light beam's profile using optical interferometry in the high angular resolution regime.

In interferometry and holography systems, the optical path difference (OPD) between the beams must be kept smaller than the source coherence length to obtain stable interference fringes. When the OPD is larger than the coherence length, the fringe phase drifts on time scales comparable to the source coherence time. This reduces fringe visibility and causes fringes to wash out completely for observation times longer than the coherence time. High fringe visibility measurements, however, can be obtained with pulsed sources (or very short detector integration times), if the pulse duration is shorter than the source coherence time, thereby limiting the amount of fringe phase drift over the pulse duration. We analyze cases where optical pulses are carved from a continuous-wave laser source. We derive an expression for the squared magnitude of the fringe visibility for pulsed interference measurements and demonstrate that high visibility fringes can be obtained even when the OPD is much longer than the source coherence length. As a rule of thumb, the source coherence time should be at least 10 times greater than the pulse duration to obtain expected fringe visibilities greater than 95% when the source laser has a Lorentzian line shape. This result is important to the design of digital holography systems for long-range remote sensing applications, where it is difficult to implement a dynamic optical delay line for path length matching with non-cooperative objects.

The detection of atmospheric carbon dioxide concentrations typically occurs within the near-infrared spectral band. To achieve high-precision measurements, it is essential not only to maintain high spectral resolution but also to capture spectral information across multiple wavelength ranges. Therefore, we propose a novel, to our knowledge, optical design for a high-resolution dual-band triple-grating spectrometer (DTGS). The design strategy is centered on fully utilizing the grating through simultaneous ±1st-order diffraction, enabling high-resolution acquisition of dual-band spectral information while enhancing system integration. We performed theoretical derivations and simulation validations for the initial structure of the DTGS system. The simulation results indicate that the DTGS system achieves high-resolution detection (0.072-0.075 nm spectral resolution) in near-infrared dual-band CO₂ weak absorption spectra (1.556-1.576µm and 1.578-1.598µm) through a grating configuration of one 555 lines/mm and two 1100 lines/mm, overcoming the spectral resolution constraints of traditional optical systems in wide-band measurements.

Fourier synthesis is one of the foundations of physical optics. Spatial Fourier optics is a basis for understanding optical imaging, microscopy, and holography. In conventional Fourier optics, the complex spatial field distribution in the Fourier plane constitutes the spatial spectrum of the field to be realized in physical space. Analogously, in temporal Fourier optics, the complex temporal spectrum can be manipulated for ultrafast pulse-shaping. We present here a tutorial on the emerging field of spatiotemporal Fourier optics whereby the spatial and temporal spectra are manipulated jointly to produce spatiotemporally structured optical fields that display unique propagation characteristics. In this tutorial, we focus on a subset of the overall class of nonseparable spatiotemporally structured fields, namely cylindrically symmetric fields in which each radial spatial frequency is associated with a single wavelength. This subset of fields comprises propagation-invariant wave packets that travel rigidly in linear media at a tunable group velocity and includes space-time wave packets and other closely related structured fields. We describe a spatiotemporal Fourier synthesis system capable of preparing arbitrary optical fields belonging to this subclass.

This study presents a class of two-dimensional (2D) spatial-frequency-modulated structures with transmittance d₁=0.10mm, in which the periodicity can vary along both the d₂=0.30mm- and n_c=1-axes. Specifically, the structure exhibits spatial frequencies n_av=3 and z=0 that sinusoidally alternate between two values along both directions, with the possibility of unequal modulation in the T(x,y)- and x-axes. It is shown that y generally behaves as an almost periodic function, resulting in an impulsive spatial spectrum. However, we identify the conditions under which f_x becomes periodic, and its spatial spectrum forms a lattice of impulses. When these periodicity conditions are met, we refer to the structure as a 2D spatially chirped periodic structure. These structures are characterized by four natural numbers, denoted as f_y, x, y, and T(x,y), which represent the modulation in the T(x,y)- and n_cx-directions, respectively, and two real parameters, named frequency modulation strengths in both the n_cy- and n_avx-directions, denoted by n_avy and x, respectively. As a special case, we define a 2D spatially chirped amplitude sinusoidal structure (SCASS), based on the transmission function of a conventional 2D amplitude sinusoidal grating, where the phase of the conventional grating is replaced by a desired chirped phase. The near-field diffraction from 2D SCASSs is studied using the angular (spatial) spectrum method. The Talbot distances for these gratings are determined and verified experimentally, showing that the intensity profiles at specific Talbot distances are highly dependent on the parameters y, x, y, k_x, k_y, and n_cx. Furthermore, we formulated the near-field diffraction of a plane wave from 2D multiplicatively separable spatially chirped amplitude sinusoidal structures, considering the variability of spatial periods in both the n_cy- and n_avx-directions. In comparison with conventional 2D gratings, new, to our knowledge, and intriguing diffraction patterns are observed, such as sharp and smooth Gaussian-like intensity spots generated via the diffraction of the incident wave, with nearly diffraction-limited features but limited overall efficiency. These intensity spots depend on the characteristic parameters of the structure. By carefully manipulating the n_avy parameters, we have the ability to generate maximum intensity peaks within these 2D SCASSs, which are

A semi-analytic theory of multilayer dielectric gratings (MLDGs) is presented. Analytic formulas for the -1st-order diffraction efficiency of an MLDG being 100% and greater than a preset value are given in terms of the scattering matrix elements of the top surface-relief grating (TG). The important role of the combined reflection phases is highlighted. The need to secure high reflectances of the multilayer stack at the two angles of incidence below the TG and the need to supply a sufficient number of aperiodic layers in the multilayer substrate for phase matching are emphasized. Three numerical examples of 100% efficient MLDGs, obtained without optimizing the TG, are presented, illustrating the accuracy and generality of the theory.

Spectroscopic optical coherence tomography enables an accurate estimation of scatterer size by computing the correlation distance (CD) function. For calibration and accuracy verification, polystyrene spheres are commonly used as size standards. However, anomalies have been observed when using the CD function to analyze spherical scatterers, which we link to multiple scattering. We have developed a robust, automated algorithm to calculate the size of hundreds of scatterers within a volumetric OCT image while accounting for the effects of multiple scattering. We measured 5.1, 7.7, and 11.3 µm polystyrene beads suspended in collagen hydrogel, agarose, and polydimethylsiloxane, which resulted in average diameters in agreement with the theoretical size to within ±λ/2, using this analysis approach. Accurately accounting for these multiple scattering effects is crucial for a robust calibration, and these measurements point the way toward analyzing the nuclear size of cells throughout a 3D tissue volume.

Fifty years after its introduction, the perturbative approach conceived by Lax et al. [Phys. Rev. A11, 1365 (1975)10.1103/PhysRevA.11.1365] to solve Maxwell's equations remains widely used, despite its inherent divergence in the physical space. Through a vectorial analysis of the LLM algorithm carried out within the spatial frequency domain, it is here shown how each single term of the asymptotic expansion of a free-space propagated coherent light field represents the outcome of a suitable linear filter fed by the boundary field. In this way, since the Fourier transform operator naturally uncouples the roles of the boundary condition and of the propagation process, the decoding mechanism of the LLM scheme can be clearly unveiled. To this end, our main task is the exploration of the convergence features of the so-called spectral series. This is the spectral counterpart of the asymptotic series representation (in the physical space) provided by the LLM algorithm. Through a delicate asymptotic analysis, it is proved that the spectral series is uniformly convergent, within the homogeneous portion of the angular spectrum, to the Fourier transform of the coherent free-space propagator. On employing Borel summation together with analytical continuation, such a fundamental connection is then extended to the remaining evanescent part of the spectrum. In this way, it is rigorously proven that (i) exact free-space solutions of Maxwell's equations are completely encoded within their paraxial approximations and (ii) the Rayleigh-Sommerfeld coherent propagator is the generating function for all nonparaxial corrections iteratively obtained by the LLM algorithm, irrespective of the boundary condition.

The flexibility and versatility of nano-assembled plasmonic structures provide platforms for mesoscale tunable optical modulation. Our recently developed model for these nano-assembled plasmonic structures is composed of a dielectric spherical core surrounded by a concentric spherical shell containing a random distribution of gold nanoparticles (AuNPs). This model provides a useful platform for studying the role of a controlled amount of disorder on scattering by a particle. In that context, we explore the angular distribution of scattered light for different sizes (5-20 nm) and filling fractions (0.1-0.3) of the AuNPs in the coatings. The simulations reveal that the coating of AuNPs redistributes power in a way that suppresses angular side lobes, thereby guiding the scattered power preferentially in the forward direction. These results highlight that, with the ability to tune both the spatial and spectral aspects of the scattering profile, these coated structures may serve as a platform for a variety of applications, including passive cloaking and high-resolution imaging.