Improving the reliability of deep learning computational ghost imaging with prediction uncertainty based on neighborhood feature maps.

Defect inspection is required in various fields, and many researchers have attempted deep-learning algorithms for inspections. Deep-learning algorithms have advantages in terms of accuracy and measurement time; however, the reliability of deep-learning outputs is problematic in precision measurements. This study demonstrates that iterative estimation using neighboring feature maps can evaluate the uncertainty of the outputs and shows that unconfident error predictions have higher uncertainties. In ghost imaging using deep learning, the experimental results show that removing outputs with higher uncertainties improves the accuracy by approximately 15.7%.

Lensless single-fiber ghost imaging.

We demonstrate lensless single-fiber ghost imaging, which allows illumination and collection using a single optical fiber without a transmission-type system. Speckle patterns with relative coincidence degrees of 0.14 were formed by image reconstruction using improved differential ghost imaging. Employing fiber with a diameter of 105 µm, we achieved a spatial resolution of 0.05 mm in an observing area of 9m m 2, at a working distance of 10 mm. Compared to a conventional neuroendoscope at a power density of 94m W/c m 2, our imaging could be realized by extremely weak illumination at a laser power density of 0.10m W/c m 2. Using our lensless single-fiber ghost imaging, with 30,000 speckle patterns and implementing a diffuser, we attained an average coincidence degree of 0.45.

Relationship between the kernel size of a convolutional layer and the optical point spread function in ghost imaging using deep learning for identifying defect locations.

We explore the contribution of convolutional neural networks to correcting for the effect of the point spread function (PSF) of the optics when applying ghost imaging (GI) combined with deep learning to identify defect positions in materials. GI can be accelerated by combining GI and deep learning. However, no method has been established for determining the relevant model parameters. A simple model with different kernel sizes was built. Its accuracy was evaluated for data containing the effects of different PSFs. Numerical analysis and empirical experiments demonstrate that the accuracy of defect identification improved by matching the kernel size with the PSF of the optics.

Noise-robust deep learning ghost imaging using a non-overlapping pattern for defect position mapping.

Defect detection requires highly sensitive and robust inspection methods. This study shows that non-overlapping illumination patterns can improve the noise robustness of deep learning ghost imaging (DLGI) without modifying the convolutional neural network (CNN). Ghost imaging (GI) can be accelerated by combining GI and deep learning. However, the robustness of DLGI decreases in exchange for higher speed. Using non-overlapping patterns can decrease the noise effects in the input data to the CNN. This study evaluates the DLGI robustness by using non-overlapping patterns generated based on binary notation. The results show that non-overlapping patterns improve the position accuracy by up to 51%, enabling the detection of defect positions with higher accuracy in noisy environments.

Fabrication of three-dimensional high-aspect-ratio structures by oblique-incidence Talbot lithography.

Developing a suitable production method for three-dimensional periodic nanostructures with high aspect ratios is a subject of growing interest. For mass production, Talbot lithography offers many advantages. However, one disadvantage is that the minimum period of the light intensity distribution is limited by the period of the diffraction grating used. To enhance the aspect ratio of fabricated nanostructures, in the present study we focus on multiple wave interference between diffracted waves created using the Talbot effect. We propose a unique exposure method to generate multiple wave interference between adjacent diffraction orders by controlling the angle of incidence of an ultraviolet (UV) light source. Using finite-difference time-domain simulations, we obtain fringe patterns with a sub-wavelength period using a one-dimensional periodic grating mask. Moreover, we demonstrate the practical application of this approach by using UV lithography to fabricate sub-wavelength periodic photopolymer-based structures with an aspect ratio of 30 in millimeter-scale areas, indicating its suitability for mass production.

Refractive index sensing with temperature compensation by a multimode-interference fiber-based optical frequency comb sensing cavity.

We proposed a refractive index (RI) sensing method with temperature compensation by using an optical frequency comb (OFC) sensing cavity including a multimode-interference (MMI) fiber, namely, the MMI-OFC sensing cavity. The MMI-OFC sensing cavity enables simultaneous measurement of material-dependent RI and sample temperature by decoding from the comb spacing frequency shift and the wavelength shift of the OFC. We realized the simultaneous and continuous measurement of RI-related concentration of a liquid sample and its temperature with precisions of 1.6 × 10-4 RIU and 0.08 °C. The proposed method would be a useful means for the various applications based on RI sensing.

Visualization of internal structure and internal stress in visibly opaque objects using full-field phase-shifting terahertz digital holography.

We construct a full-field phase-shifting terahertz digital holography (PS-THz-DH) system by use of a THz quantum cascade laser and an uncooled, 2D micro-bolometer array. The PS-THz-DH enables us to separate the necessary diffraction-order image from unnecessary diffraction-order images without the need for spatial Fourier filtering, leading to suppress the decrease of spatial resolution. 3D shape of a visibly opaque object is visualized with a sub-millimeter lateral resolution and a sub-µm axial resolution. Also, the digital focusing of amplitude image enables the visualization of internal structure with the millimeter-order axial selectivity. Furthermore, the internal stress distribution of an externally compressed object is visualized from the phase image. The demonstrated results imply a possibility for non-destructive inspection of visibly opaque non-metal materials.

Strain sensing based on strain to radio-frequency conversion of optical frequency comb.

We propose an optical frequency comb (OFC)-based strain sensing method, namely OFC sensing cavity, which is capable of radio-frequency (RF)-based strain measurement. We developed a null-method-based strain sensing system with a comb-spacing-stabilized OFC generator. We realized strain measurement from 1.83 µÎµ to 1800 µÎµ with a sensing fiber length of 20 mm. The measurable strain frequency range of the developed strain sensing system was from 0 to 310 Hz. Owing to the use of RF-based strain measurement, our approach would be a useful and powerful tool for sensing of strain or other physical quantities, and the concept of the OFC sensing cavity is a new aspect of OFC technology.

Refractive-index-sensing optical comb based on photonic radio-frequency conversion with intracavity multi-mode interference fiber sensor.

Optical frequency combs (OFCs) have attracted attention as optical frequency rulers due to their tooth-like discrete spectra together with their inherent mode-locking nature and phase-locking control to a frequency standard. Based on this concept, their applications until now have been demonstrated in the fields of optical frequency metrology. However, if the utility of OFCs can be further expanded beyond their application by exploiting new aspects of OFCs, this will lead to new developments in optical metrology and instrumentation. Here, we report a fiber sensing application of OFCs based on a coherent link between the optical and radio frequencies, enabling high-precision refractive index measurement based on frequency measurement in radio-frequency (RF) region. Our technique encodes a refractive index change of a liquid sample into a repetition frequency of OFC by a combination of an intracavity multi-mode-interference fiber sensor and wavelength dispersion of a cavity fiber. Then, the change in refractive index is read out by measuring the repetition frequency in RF region based on a frequency standard. Use of an OFC as a photonic RF converter will lead to the development of new applications in high-precision fiber sensing with the help of functional fiber sensors and precise RF measurement.

Multicascade-linked synthetic wavelength digital holography using an optical-comb-referenced frequency synthesizer.

Digital holography (DH) is a promising method for non-contact surface topography because the reconstructed phase image can visualize the nanometer unevenness in a sample. However, the axial range of this method is limited to the range of the optical wavelength due to the phase wrapping ambiguity. Although the use of two different wavelengths of light and the resulting synthetic wavelength, i.e., synthetic wavelength DH, can expand the axial range up to several hundreds of millimeters, its axial precision does not reach sub-micrometer. In this article, we constructed a tunable external cavity laser diode phase-locked to an optical frequency comb, namely, an optical-comb-referenced frequency synthesizer, enabling us to generate multiple synthetic wavelengths within the range of 32 µm to 1.20 m. A multiple cascade link of the phase images among an optical wavelength ( = 1.520 µm) and 5 different synthetic wavelengths ( = 32.39 µm, 99.98 µm, 400.0 µm, 1003 µm, and 4021 µm) enables the shape measurement of a reflective millimeter-sized stepped surface with the axial resolution of 34 nm. The axial dynamic range, defined as the ratio of the axial range ( = 2.0 mm) to the axial resolution ( = 34 nm), achieves 5.9 × 105, which is larger than that of previous synthetic wavelength DH. Such a wide axial dynamic range capability will further expand the application field of DH for large objects with meter dimensions.

Scan-less hyperspectral dual-comb single-pixel-imaging in both amplitude and phase.

We have developed a hyperspectral imaging scheme that involves a combination of dual-comb spectroscopy and Hadamard-transform-based single-pixel imaging. The scheme enables us to obtain 12,000 hyperspectral images of amplitude and phase at a spatial resolution of 46 µm without mechanical scanning. The spectral resolution given by the data point interval in the frequency domain is 20 MHz and the comb mode interval is 100 MHz over a spectral range of 1.2 THz centered at 191.5 THz. As an initial demonstration of our scheme, we obtained spectroscopic images of a standard test chart through an etalon plate. The thickness of an absorptive chromium-coated layer on a float-glass substrate was determined to be 70 nm from the hyperspectral phase images in the near-infrared wavelength region.

Real-time absolute frequency measurement of continuous-wave terahertz radiation based on dual terahertz combs of photocarriers with different frequency spacings.

Real-time measurement of the absolute frequency of continuous-wave terahertz (CW-THz) radiation is required for characterization and frequency calibration of practical CW-THz sources. We proposed a method for real-time monitoring of the absolute frequency of CW-THz radiation involving temporally parallel, i.e., simultaneous, measurement of two pairs of beat frequencies and laser repetition frequencies based on dual THz combs of photocarriers (PC-THz combs) with different frequency spacings. To demonstrate the method, THz-comb-referenced spectrum analyzers were constructed with a dual configuration based on dual femtosecond lasers. Regardless of the presence or absence of frequency control in the PC-THz combs, a frequency precision of 10(-11) was achieved at a measurement rate of 100 Hz. Furthermore, large fluctuation of the CW-THz frequencies, crossing several modes of the PC-THz combs, was correctly monitored in real time. The proposed method will be a powerful tool for the research and development of practical CW-THz sources, and other applications.

Double-modulation reflection-type terahertz ellipsometer for measuring the thickness of a thin paint coating.

We constructed a double-modulation, reflection-type terahertz (THz) ellipsometer for precise measurement of the thickness of a paint film which is coated on a metal surface and which is not transparent to visible or mid-infrared light. The double-modulation technique enabled us to directly obtain two ellipsometric parameters, Δ(ω) and Ψ(ω), as a function of angular frequency, ω, with a single measurement while reducing flicker noise due to a pump laser. The bias voltage of a photoconductive antenna (PCA) used as a THz pulse emitter was modulated at 100 kHz, and a first lock-in amplifier (LA1) was connected to the output of an electro-optic (EO) signal-sampling unit. In addition, a wire-grid polarizer (WGP) was rotated at 100 Hz to conduct polarization modulation with a frequency of 200 Hz. The output signal from LA1 was fed into a second lock-in amplifier (LA2) that worked in synchronization with the rotating WGP (RWGP). By operating LA2 in a quadrature phase-detection mode, we were able to obtain in-phase and out-of-phase signals simultaneously, from which the two ellipsometric parameters for an isotropic sample could be derived at the same time while cancelling common-mode noise. The lower detection limit of the thickness measurement and the relative standard deviation (RSD) of a black paint film coated on an aluminum substrate were 4.3 µm and 1.4%, respectively. The possibility of determining all elements of the Jones matrix for an anisotropic material is also discussed.

Ellipsometric measurement technique for a modified Otto configuration used for observing surface-plasmon resonance.

In order to carry out precise measurements of the thickness of a dielectric layer deposited on a metal surface, we have introduced an ellipsometric measurement technique (EMT) to the modified Otto's configuration (MOC) that is used for observing surface plasmon resonance (SPR). For that purpose, we have measured the thickness of the Au layer by the EMT at four different locations on an elliptic fringe pattern obtained from the MOC basing on a four-layer structure model: prism (BK7)-air-Ausubstrate (BK7). Then, we have measured that of a TiO(2) layer deposited on the Au layer by the EMT basing on a five-layer structure model: prism (BK7)-air-TiO(2)-Au-substrate (BK7). We have found experimentally that the combination of EMT and MOC is effective for measuring the thickness of the dielectric layer on the metal.

Dual-comb spectroscopic ellipsometry.

Spectroscopic ellipsometry is a means of investigating optical and dielectric material responses. Conventional spectroscopic ellipsometry is subject to trade-offs between spectral accuracy, resolution, and measurement time. Polarization modulation has afforded poor performance because of its sensitivity to mechanical vibrational noise, thermal instability, and polarization-wavelength dependency. We combine spectroscopic ellipsometry with dual-comb spectroscopy, namely, dual-comb spectroscopic ellipsometry. Dual-comb spectroscopic ellipsometry (DCSE). DCSE directly and simultaneously obtains the ellipsometric parameters of the amplitude ratio and phase difference between s-polarized and p-polarized light signals with ultra-high spectral resolution and no polarization modulation, beyond the conventional limit. Ellipsometric evaluation without polarization modulation also enhances the stability and robustness of the system. In this study, we construct a polarization-modulation-free DCSE system with a spectral resolution of up to 1.2 × 10-5 nm throughout the spectral range of 1514-1595 nm and achieved an accuracy of 38.4 nm and a precision of 3.3 nm in the measurement of thin-film samples.Spectroscopic ellipsometry is an established technique to characterize the optical properties of a material. Here, Minamikawa et al. combine the method with dual-comb spectroscopy, which allows them to obtain ellipsometric parameters including the phase difference between s-polarized and p-polarized light.

Dynamic terahertz spectroscopy of gas molecules mixed with unwanted aerosol under atmospheric pressure using fibre-based asynchronous-optical-sampling terahertz time-domain spectroscopy.

Terahertz (THz) spectroscopy is a promising method for analysing polar gas molecules mixed with unwanted aerosols due to its ability to obtain spectral fingerprints of rotational transition and immunity to aerosol scattering. In this article, dynamic THz spectroscopy of acetonitrile (CH3CN) gas was performed in the presence of smoke under the atmospheric pressure using a fibre-based, asynchronous-optical-sampling THz time-domain spectrometer. To match THz spectral signatures of gas molecules at atmospheric pressure, the spectral resolution was optimized to 1 GHz with a measurement rate of 1 Hz. The spectral overlapping of closely packed absorption lines significantly boosted the detection limit to 200 ppm when considering all the spectral contributions of the numerous absorption lines from 0.2 THz to 1 THz. Temporal changes of the CH3CN gas concentration were monitored under the smoky condition at the atmospheric pressure during volatilization of CH3CN droplets and the following diffusion of the volatilized CH3CN gas without the influence of scattering or absorption by the smoke. This system will be a powerful tool for real-time monitoring of target gases in practical applications of gas analysis in the atmospheric pressure, such as combustion processes or fire accident.

Adaptive sampling dual terahertz comb spectroscopy using dual free-running femtosecond lasers.

Terahertz (THz) dual comb spectroscopy (DCS) is a promising method for high-accuracy, high-resolution, broadband THz spectroscopy because the mode-resolved THz comb spectrum includes both broadband THz radiation and narrow-line CW-THz radiation characteristics. In addition, all frequency modes of a THz comb can be phase-locked to a microwave frequency standard, providing excellent traceability. However, the need for stabilization of dual femtosecond lasers has often hindered its wide use. To overcome this limitation, here we have demonstrated adaptive-sampling THz-DCS, allowing the use of free-running femtosecond lasers. To correct the fluctuation of the time and frequency scales caused by the laser timing jitter, an adaptive sampling clock is generated by dual THz-comb-referenced spectrum analysers and is used for a timing clock signal in a data acquisition board. The results not only indicated the successful implementation of THz-DCS with free-running lasers but also showed that this configuration outperforms standard THz-DCS with stabilized lasers due to the slight jitter remained in the stabilized lasers.

Autoregressive-model-based fluorescence-lifetime measurements by phase-modulation fluorometry using a pulsed-excitation light source and a high-gain photomultiplier tube.

We propose a novel method for measuring fluorescence lifetimes by use of a pulsed-excitation light source and an ordinary or a high-gain photomultiplier tube (PMT) with a high-load resistor. In order to obtain the values of fluorescence lifetimes, we adopt a normal data-processing procedure used in phase-modulation fluorometry. We apply an autoregressive (AR)-model-based data-analysis technique to fluorescence- and reference-response time-series data obtained from the PMT in order to derive plural values of phase differences at a repetition frequency of the pulsed-excitation light source and its harmonic ones. The connection of the high-load resistor enhances sensitivity in signal detection in a certain condition. Introduction of the AR-model-based data-analysis technique improves precision in estimating the values of fluorescence lifetimes. Depending on the value of the load resistor and that of the repetition frequency, plural values of fluorescence lifetimes are obtained at one time by utilizing the phase information of harmonic frequencies. Because the proposed measurement system is simple to construct, it might be effective when we need to know approximate values of fluorescence lifetimes readily, such as in the field of biochemistry for a screening purpose.