This PDF file contains the front matter associated with SPIE Proceedings Volume 12813, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

A denoising convolutional neural network based on batch-renormalization (BRDNet) is proposed to denoise Brillouinga in spectrum, and its performance is compared with CNN-based denoising network (DnCNN). Experimental results show that BRDNet can improve the signal-to-noise ratio (SNR) of 15dB and obtain higher measurement accuracy than DnCNN.

Event identification based on Ф-OTDR stands as a prominent research hotspot in the realm of distributed fiber optic sensing. However, the challenge of collecting rare vibration signals in Ф-OTDR applications limits the model's ability to classify accurately. We propose a TSG data augmentation method based on Time Series Transfer and StyleGAN. The augmented data is enhanced for diversity and generalization. With a sample size of 50 per class, our approach enhances the classification accuracy from 66.89% to 81.61%.

A distributed optical fiber magnetic field sensor based on polarization-sensitive optical frequency domain reflectometer (POFDR) is proposed. The sensor employs a novel distributed magnetic field measurement method that firstly extracts the magnetic field induced nonreciprocal circular birefringence by the combination of the Stokes vectors and the backward Mueller matrices from the measured state of polarization (SOP). Then, the accumulated Faraday rotation angles can be further calculated to characterize the distribution of the magnetic field. It overcomes the drawback of the conventional POFDR scheme that requires at least two different input SOPs for each sensing. Finally, the aforementioned effectiveness has been experimentally verified by using a single-mode fiber as the sensing fiber. The distribution of a static magnetic field in milliTesla has been successfully measured at a spatial resolution of 0.16 m.

This paper reports a noise suppression technique for Φ-OTDR based on Rayleigh scattering enhanced points. Using femtosecond laser to write scattering-enhanced points (SEPs) in the sensing fiber of Φ-OTDR, each SEP is composed of three points spaced 4mm apart. Experimental results show that the SNR of Φ-OTDR based on scattering-enhanced fiber is improved by 24 dB, and the system phase noise is better than -80 dB (re: 1 rad/√Hz), the three-point structure has better stability and higher reinforcement effect. Compared with the unenhanced Φ-OTDR system, the noise is reduced by about 17 dB.

Indium tin oxide (In₂O₃-SnO₂-90/10 wt%, ITO) is a semiconductor material with excellent electrical conductivity. In this paper, ITO was deposited on a multimode optical fiber by magnetron sputtering technique and characterized by using Scanning Electron Microscopy (SEM). It is subsequently used as a working electrode in a three-electrode system to study its electrochemical behavior in different solutions. In 0.1 M KCl containing such redox probes as 1 mMofK₃[Fe(CN)₆] were discussed by Cyclic voltammetry (CV) method at different scan rates. The observed electrochemical processes are quasi-reversible and diffusion-controlled. The results of the investigation inject new vitality to enhance the intersection of electrochemistry and optics disciplines, and also lay the foundation for dual-domain determination.

A dynamic strain range extension method is proposed for fiber distributed sensing system based on dual-sideband frequency modulation pulse. Aiming at the problem of huge raw data and slow processing speed of the dual-sideband system, the RF circuit module scheme and corresponding algorithms are proposed to enhance the system's ability to apply to real-world scenarios. This study extends the range of measurable dynamic strain in the system and effectively tackles the challenges of storage and computational efficiency as data volume increases. It enhances the system's ability to adapt to complex environments.

We propose and demonstrate an optical fiber probe which can simultaneously detect low and high-wavenumber coherent anti-Stokes Raman scattering (CARS) spectrum of samples. The low and high-wavenumber resonance signals are excited by the pump pulses and dual-Stokes pulses which are generated by the soliton self-frequency shift. The optical fiber taper probe focuses and delivers CARS excitation pulses to the samples. The detection of low and high-wavenumber regions can be achieved by the characterization of CN triple stretching vibrations and CH stretching vibrations. The simultaneous low and high-wavenumber optical fiber CARS probe will enable wider applications of quantitative chemical detection in in vivo biomedical research.

Phase interrogation method can effectively avoid the operational spacing problems of fiber-optic Fabry-Perot (F-P)sensors. Based on the principles of F-P interference and low coherence interference, polarizers and birefringent crystals are applied to construct signals with quadrature relationship. A four-quadrant inverse tangent operation is employed to accurately calculate the phase values. We have performed rapid, high-speed measurements of dynamic pressure with aF-P sensor. Experimental results show that the method can achieve real-time pressure measurements up to 3 MPa with an interrogation rate of 5 kHz. This research holds much promise for the promotion of interferometric based fiber-optic sensors and applications of pressure measurements.

We have theoretically and experimentally demonstrated the effectiveness of a modulated asymmetric dual Mach-Zehnder interferometer (ADMZI) system using a high-frequency carrier scheme, which is also applicable to weak signals with low frequencies and small amplitudes. To further extend the sensing distance, we choose an erbium-doped fiber amplifier (EDFA) post-amplification to amplify the interference signal. The experimental results show that the scheme extends the sensing distance to 200 km, and the positioning error is within ±300 m with a standard deviation of 72.2 m. The average positioning time is 195 ms, which can meet the real-time positioning requirements.

A difference-frequency coherent optical time domain reflectometry (DF-COTDR) technique is proposed for distributed acoustic sensing (DAS) systems with coherent detection and ultra-low sampling rate. In this research, we first analyze the power spectrum of the beat signal and propose a method of using local oscillator (LO) light to modulate the frequency to reduce the requirement for high sampling rate. The results show that the corresponding linearity of DFCOTDR reaches 99%. And the demodulation duration is shortened by more than four times. In addition, this method can be generalized to existing COTDR systems with a few modifications.

In this paper, a method of measuring NH₃ gas concentration based on fuzzy neural network is proposed. The 1560nmlaser is used as the pump laser source, and an all-fiber thulium-doped fiber ring laser system is built. Wave length scanning technology is used to scan the spectrum of different concentrations of NH₃ gas in the range of 1957-1960nm, and the spectral data is collected and preliminary-processed. Then the fuzzy neural network model is used to measure the concentration of NH3 gas. The method proposed in this paper can effectively avoid the influence of temperature and pressure changes and measurement errors of experimental equipment on ammonia concentration detection. The experimental results show that the system has high accuracy and strong robustness.

The resonant enhancement of both mechanical and optical response in microcavity optomechanical devices allows exquisitely sensitive measurements of stimuli in environment. We demonstrate and characterize a optomechanical oscillator based on a hollow microbubble and employ it as a sensor for acoustic wave. The output spectrum has a unique waveform that consists of fast and slow oscillation periods. The sensitivity of acoustic sensing has been greatly improved and the noise equivalent sound pressure as low as 0.89 Pa is obtained.

A novel vibration detection system based on directly modulated chirped pulses and a grating pair is proposed. Information of the vibration is obtained by demodulating the phase change of the interference fringes from reflection of the grating pair. The system has an operating frequency range of 1 Hz to 30 kHz, and the difference in frequency response between the system and a commercial accelerometer is less than 1%. This vibration detection system provides a sensitive and high-speed vibration detection scheme with a wide frequency range. At the same time, it has the advantages of low cost and simple structure, making it an attractive option for various applications.

A temperature-insensitive strain sensor based on a microwave photonic filter (MPF) is proposed. A Mickelson interferometer, formed by two identical chirped fiber Bragg gratings (FBGs), together with a broadband optical source, a phase modulator, a dispersion compensation fiber, and a photodetector, forms the MPF. Due to the dispersion effect of chirped FBG, the wavelength shift caused by strain will result in changes in signal transmission latency, which will then cause a drift on the central frequency of the MPF. An experiment was conducted to validate the proposed scheme. Experimental results show that frequency shift of MPF in response to strain with a sensitivity of 2.0 GHz is obtained for a measurement range of 0~1.8 mε. Besides, the system keeps stable within a temperature range from 30 °C to 100 °C.

Sensitive detection of CO₂ concentrations is exceedingly significant for ecological environmental protection and people's production safety. The convolutional layer of convolutional neural network (CNN) can automatically extract features, which greatly saves the cost of manual processing. In this paper, we build a gas sensing system based on thulium-doped ring laser to obtain CO₂ absorption spectral data. A one-dimensional CNN model was proposed to process the data and achieve accurate prediction of CO₂ concentrations. The prediction of the test set data resulted in a regression coefficient R² of 99.58% and MSE of 0.010, which meets the gas detection requirements. The combination of deep learning algorithm and gas absorption spectroscopy provides new ideas for absorption spectroscopy-based gas sensing technology.

This paper is to introduce a high-robustness polarization-based optical fiber Fabry-Perot microphone system. This system features an extrinsic Fabry-Perot interferometric (EFPI) sensor with a polyphenylene sulfide (PPS) diaphragm for signal detection. It is further integrated into a cross-correlation interrogation system using polarization low-coherence interference technology with a birefringent crystal. The interrogation system does not require an orthogonal relationship between the interference signals. Moreover, it can still work when the initial cavity length of the sensor drifts due to the influence of the environment. The experiments have been carried out to verify broadband signal and voice detection capability. The experimental results showcase several notable advantages of the proposed system, including a wide frequency range, high sensitivity. Consequently, the system holds tremendous potential for diverse practical engineering applications.

A novel technology called visual robotic arm aided laser standoff methane sensing is proposed in this paper for intelligent inspection of gas pipeline leakage. It mounts a homemade miniaturized laser methane sensor to the end of a robotic arm and uses a depth camera to guide the laser towards the target pipeline. By improving Yolov5s, real-time high-precision identification of pipelines was achieved with an accuracy rate of 99.5%. Collaborative work is achieved through hand eye calibration. The system able to locate leaks at a distance of 3m with an error of approximately 30mm.