Optics express最新文献

We present an all-silicon anisotropic effective-medium half-wave plate (HWP) that provides broadband, low-loss polarization control in the terahertz (THz) band. A two-dimensional anisotropic effective medium was created by etching elliptical air holes on a rhombic lattice through high-resistivity silicon using standard techniques based on micro-electromechanical systems. Finite element simulations predict a phase difference between orthogonal polarizations of π ± 0.08π across 1.2-2.0 THz for a single 200 µm-thick layer. The results of terahertz time-domain spectroscopy are presented to demonstrate this HWP performance. Moreover, they show that stacking two, three, and four identical layers to effective thicknesses of 400, 600, and 800 µm shifted the operating band to 0.6-1.0, 0.4-0.67, and 0.3-0.5 THz, respectively, without degrading the phase response. Experimental spectra agree closely with the theory, showing that arbitrary effective thicknesses-and thus target frequency windows-can be achieved by simply stacking thin, easily processed wafers. Parametric sweeps indicated that performance was preserved under standard photolithography tolerances and lateral misalignments of up to 10% of the lattice period. By combining high transmission, wide bandwidth, and straightforward fabrication, the proposed device offers a scalable route for compact THz-band polarization optics for spectroscopy, imaging, and next-generation (6G) wireless communications.

In high dynamic range (HDR) imaging, exposure parameters determine the efficiency of image acquisition and have a great impact on image quality. Two exposure selection methods are proposed to improve the efficiency of HDR imaging while maintaining image quality. The target signal-to-noise ratio (SNR) is derived from the contrast sensitivity of the human visual system under a given adaptation condition. Combining the target SNR and imaging model, a histogram-independent iterative method and a histogram-based exhaustive method are developed to determine the exposure time set. The two methods are complementary: the former offers low computational cost, while the latter enhances control over exposure parameters and improves image acquisition efficiency. An HDR dataset with 1/3-stop exposure intervals and diverse illumination conditions is established, which is utilized alongside a publicly available HDR dataset to evaluate the proposed methods in comparison with existing methods. The proposed methods achieve high quality while reducing image capture time, demonstrating the superior efficiency of HDR image acquisition.

In this work, we propose a methodology for grain moisture content detection based on a spoof localized surface plasmons (SLSPs) resonator, enabling highly efficient and precise quantification of moisture levels in grains. We conducted tests within the 0.9-1.1 GHz frequency range on nine wheat samples with varying moisture contents, successfully obtaining corresponding transmission spectra. Through calculation and in-depth analysis, a linear regression equation for wheat moisture content was established, with a coefficient of determination (R²) as high as 0.985. Experimental data indicated that the root mean square error (RMSE), mean absolute error (MAE), and maximum relative error (MRE) of predicted values were 0.321%, 0.24%, and 7.6%, respectively.

The comb-like spectrum added to laser light by an electro-optic modulator (EOM) finds use in a wide range of applications, including coherent optical communication, atomic spectroscopy, and laser frequency and phase stabilization. In some cases a sideband-free optical frequency shift is preferred, such as in laser offset locking using an optical cavity, single-photon frequency shifting, and laser range finding. Approaches to obtaining an optical frequency offset (OFO) involve trade-offs between shift range, conversion gain, and suppression of spurious sidebands. Here we demonstrate an OFO of continuous-wave 871 nm laser light by serrodyne modulation using a fiber EOM and radio-frequency (RF) tones from a commercial RF system on a chip (RFSoC) to achieve shifts of 40 to 800 MHz with >15 dB suppression of spurious sidebands and <1.5 dB conversion loss. We also observe a smoothly varying conversion gain. The utility of this tool is demonstrated by continuously shifting the offset of a cavity-locked laser from 50 to 1600 MHz, a capability useful in spectroscopy of unknown optical transitions.

Spectroscopic studies on Yb³⁺ and Dy³⁺ co-doped, and Yb³⁺, Er³⁺, and Dy³⁺ triple-doped ZBLAN (ZrF₄-BaF₂-LaF₃-AlF₃-NaF) glasses are reported. Efficient energy transfers from Yb³⁺ directly to Dy³⁺, and from Yb³⁺ to Er³⁺ and then from Er³⁺ to Dy³⁺ are demonstrated, and their rates are calculated from the measured lifetimes. This discovery enables the development of high-power Dy³⁺ doped fiber lasers pumped by low-cost, high-efficiency InGaAs diodes within the ytterbium absorption band.

To address the issue of insufficient imaging contrast in strong backlight environments, this paper proposes a dynamic light field modulation method based on liquid crystal spatial light modulator (LC-SLM) and dual-polarization system. This system constructs a cross-scale polarization filtering coupling model. By precisely controlling the transmittance of LC-SLM with voltage and dynamically adjusting the angle between the dual-polarization plates, the incident light field can be dynamically regulated. Experimental results show that in three different brightness conditions, the method proposed in this paper significantly outperforms traditional imaging and dual-polarization imaging methods in terms of image contrast and detail resolution. Specifically, the contrast of intensity imaging is respectively improved by an average of 51%, 31%, and 17% compared to traditional methods, and the contrast of polarization imaging is respectively improved by an average of 85%, 83%, and 62%. In terms of target detail resolution, intensity imaging is respectively improved by an average of 278%, 452%, and 309% compared to traditional methods, and polarization imaging is respectively improved by an average of 153%, 606%, and 568%. This research provides an innovative technical approach for solving the problem of high-contrast imaging in strong backlight environments and has important application value in intelligent driving, remote sensing mapping, and other fields.

Accurate, real-time inversion of the plume extinction coefficient is of great significance in remote sensing, environmental protection, and health security. This paper proposes the dual iterative ratio (DIR) method for retrieving the plume extinction coefficient, based on the ratio of the Mie scattering echo signal to the background signal with adaptive iterative optimization. Simulation tests demonstrate that the DIR method performs well under low signal-to-noise ratios (SNR) and is compatible with varying plume concentrations and lidar ratios; when the SNR exceeds 10, the average relative error is Less than 0.343. The algorithm is insensitive to changes in the geometric factor caused by optical misalignment: with a 1.0 mrad boresight offset in a coaxial transceiver lidar, the average relative error changes by only 4.55%. In experiments, the DIR method exhibits a strong correlation with classical inversion algorithms and achieves an average relative error of 0.131, outperforming classical methods. The measured average computation time is 80.8ms, making it suitable for real-time detection systems. The proposed algorithm provides a convenient and accurate approach of obtaining the plume extinction coefficient, addressing a technical gap in the field.

We present a unified and transparent theoretical framework for two longstanding phenomena in light-matter systems: collective emission and anomalous transport. Although both subjects have been extensively studied, we revisit them using a field-theoretical approach that solves the dressed Bloch wave functions of photons and electrons in an irradiated array of two-level atoms. This treatment reveals that the evolution of emission with irradiation frequency-from subradiance to superradiance-is directly encoded in the dressed photon momentum, which captures radiative decay and coherence. At the same time, photon-mediated interactions give rise to unconventional transport responses, including near-zero and even negative resistivity, whose microscopic origin can be consistently explained through the self-energy of the dressed electron. Framed in terms of Bloch polaritons, our results establish a direct link between radiative decay, cooperative interference, and transport anomalies, providing fresh insights into engineered light-matter platforms with tunable emission and transport properties.

While the nonlinear Shrödinger equation (NLSE) and its solving via the split-step Fourier method are well established when studying the Kerr interactions in waveguides, it is typically not applied when modeling a nonlinear interaction in a Bragg grating (BG). In that specific case, the solving of a set of coupled equations is preferred as they form the natural framework to deal with co- and contra-propagating waves. This, however, has limitations for input spectra much larger than this bandgap, e.g., for frequency combs or multispectral pump schemes. In order to deal with those in a Bragg grating, we adapt the usual NLSE solving via split-step Fourier by embedding the Bragg resonance into the dispersion operator. Although it requires that the total nonlinearity along the propagation remains moderate, i.e., the nonlinear phase shift γPL < 2π, and the pump(s) frequency(ies) to be outside of the bandgap, this modeling allows us to retrieve established results and points towards the BG ability to tune and quench four-wave mixing processes.

Most existing decryption methods for chaotic optical communication rely on chaos synchronization, which can be susceptible to external interference and suffer from performance limitations. This paper proposes a chaotic optical communication decryption framework based on a cascaded feedforward neural network (CFNN). The framework constructs a two-dimensional matrix corresponding to intermediate features within the neural network using BiMatch. Through the continuous inference of the cascaded neural network, it progressively extracts the features of the encrypted signal and ultimately recovers the message, without the need for chaos synchronization, alignment, and differential. Under the condition of maintaining comparable parameter size and computational complexity to traditional models, CFNN can reduce the bit error rate (BER) to below 3.8 × 10^-3 in most cases, demonstrating clear advantages in both decryption accuracy and robustness. Additionally, security analysis and experimental validation further confirm the potential for the practical application of the proposed method.