Based on the basic theory of vector radiative transfer equation, considering the influence of different kinds of cloud layers and different concentrations of advection fog, we developed a simplified two-layer atmosphere structure to a three-layer one and applied an adding-doubling method to simulate the scattering process of skylights through multilayer atmosphere. Furthermore, with the Rayleigh and Mie theory, the distribution characteristics for atmosphere, cloud, and fog layers were described, respectively, and the adding-doubling method was utilized to build the polarization distribution model of skylights under different typical marine environments. In addition, we investigated the relationship between the degree of polarization (DOP) and the observed altitude angle on the meridian of the sun. The results show that the skylight DOP varied regularly with the observed altitude angle. Moreover, with the rise in the optical thickness and with the increase of the cloud size and advection fog particles, DOP gradually increased. In conclusion, this study provides theoretical references for the selection of weather conditions and observation locations in the vertical polarization detection under complicated marine environments.

The stimulated Brillouin scattered (SBS) light caused by laser irradiation at target is modulated when passing through a continuous phase plate (CPP), which causes damage to optical elements. In this paper, a theoretical model for analyzing back near-field diffraction transmission of SBS light is established, which is combined with Fresnel diffraction theory and used to analyze and numerically simulate the influences of diffraction distance, CPP parameters and application position on near-field diffraction modulation. The results show that as the diffraction distance increases within a certain range, the near-field modulation increases rapidly. Near-field modulation is affected by CPP phase amplitude, minimum spatial period and spectral control, in which phase amplitude has a relatively high impact. In addition, the effect of doubling frequency CPP on near-field modulation is weaker than that of fundamental frequency CPP. The study of back near-field diffraction transmission of CPP is helpful for understanding the near-field intensity distribution of a phase optical device and provides some guidance on the arrangement of optical elements of a high-power laser system and the optimization of CPP.

Based on Fresnel scalar diffraction theory, the analytical formula of far-field spot distribution of laser emission system in defocused state is derived, and a sampling interval optimization method based on similarity coefficient is proposed in this paper. The deviation between the analytical results from optimal sampling interval and numerical simulation is less than 0.04%. Fourier transform method is used to simulate the far-field spot distribution in the presence of aberration, and the influence of key parameters such as angular displacement, divergence angle, and gain on the light field distribution is analyzed. When the laser wavelength is λ, the allowable pointing deviation is 3 μrad, and the center gain margin is 3 dB, based on the tolerance allocation theory, the tolerances of defocus, tilt, spherical aberration, coma aberration, and astigmatism should be less than or equal to 0.36λ, 0.07λ, 0.43λ, 0.13λ, and 0.5λ, respectively.

First, the spectral characteristics of multi-phase-shifted fiber Bragg gratings (MPSFBGs) were studied theoretically. The results confirmed that we could acquire the spectra with narrowband flat-top filter response by accurately designing the key parameters of the MPSFBGs, such as the number and position distribution of phase shifts, the coupling coefficient and length of the gratings. However, the process errors introduced in the inscription process would worsen the filter spectra. To this end, MPSFBGs were inscribed by a post-ultraviolet-radiation technique and the relevant spectra were measured. The results demonstrate that the experimental spectra are in agreement with the ideal simulation results when the MPSFBGs have a large bandwidth, indicating that the control accuracy can meet the requirements of the gratings. However, the performance of the narrow-bandwidth filter with a larger coupling coefficient is more sensitive to the process errors. Furthermore, the process errors were analyzed and discussed. It turns out that increasing the length of MPSFBGs can effectively widen the tolerance of phase-shift position distribution, and phase-shift difference, and weaken the effect of UV-induced loss on the loss characteristic of the filters.

Based on the form of the microlens array, a new type of large-field laser communication receiving optical system is designed, and a 3×3 optical matrix model that completely describes the light transmission in the microlens array is proposed. The influences of the tilt angle and lateral shift of each optical element on the image plane height and exit angle are discussed. According to the design requirements of the microlens array optical system, reasonable tolerance ranges of the tilt angle and lateral shift are proposed. Based on the discussion of the aberration of the integral lens optical system, the combining of the optical design and the optical simulation can be used to realize the design of a large field of view laser communication optical system. The correctness of the 3D optical matrix model is also verified. Through the prototype development, uniform light test, and field of view test, a new type of laser communication receiving optical system with field of view of 0.9° and uniformity of 86.58% is finally realized. The results show that the experimental test data is consistent with the theoretical simulation data. The feasibility and superiority of the application of the microlens array optical system in a laser communication system are further proved by the discussion and analysis on the laser communication link, which provides a new idea and new direction for the design and development of laser communication receiving optical systems.

Meijer G function is used to derive asymptotic expressions of average bit error rate and outage probability of adaptive M-ary phase-shift keying (MPSK) modulation system based on joint probability density function of misaligned fading (i.e., pointing error) and Malaga turbulence distribution. When receiver radius is fixed, the optimal bit error rate can be obtained by adjusting the beam waist radius and jitter standard deviation. When the ratio of the beam waist radius to the receiver radius is wz/a=10 and jitter standard deviation to the receiver radius is σs/a=1.5, the average bit error rate curve of non-adaptive binary phase-shift keying (BPSK) system drops the fastest. The average bit error rate and spectral efficiency of the adaptive MPSK system are both optimal when wz/a=5 and σs/a=1.5. When the misaligned fading is constant and average electric signal-to-noise ratio is small, the average bit error rate and outage probability of adaptive system are mainly affected by the misaligned fading, but with the increase in the average electric signal-to-noise ratio, the influence of turbulence increases.

The artifacts of radial edges are obvious shortcoming of external CT image. In order to suppress the artifacts of the radial edge of external CT images the weighted directional total variation (WDTV) algorithm is used to calculate the local directional differences along the radial and tangential directions, and two weighted parameters are introduced to carry out the weighted sum of these two local directional differences. The WDTV algorithm can better describe the sparsity of external CT image gradient magnitude and effectively improve the quality of reconstructed image. In order to meet the needs of high-noise neutron external CT inspection, different total variation (TV) minimization constraints are applied in certain angle ranges near the radial and tangential directions under the framework of WDTV reconstruction model. In the improved WDTV algorithm, TV minimization plays a more obvious role and has stronger performance of suppressing radial edge artifacts and anti-noise. The results of computer simulation and gear cold neutron experiment show that the improved WDTV reconstruction model can suppress the noise of reconstructed images more effectively and improve the quality of reconstructed images.

Low-dose medical computed tomography (CT) images are associated with noise problems, and it is difficult to obtain relevant paired datasets. To solve these issues, we propose a low-dose CT denoising algorithm, which is based on an improved cycle generative adversarial network. Our algorithm achieves end-to-end mapping from low-dose CT images to standard-dose CT images using unpaired datasets. In addition, to make the generator output image similar to the target image, we creatively put the DenseNet residual learning network model to the generator, wherein feature reusability is beneficial to restore the image details. Research confirms that this algorithm effectively improves the ability of edge keeping and denoising. The quality of the restored image is significantly improved, which is helpful for the detection and analysis of lesions.

Based on the analysis of the advantages and disadvantages of the existing algorithms, various types of images are considered as far as possible. Based on the principle prototype of light emitting diode-liquid crystal display direct-down image and video display, a reliable and effective data measurement method based on regional backlight brightness extraction is proposed, and an efficient and practical backlight brightness extraction method based on deep learning is proposed. The method is based on the neural network with a multi-layer down sampling structure to extract the image features and obtain the optimal backlight. The experimental results show that the method can improve the image display quality and expand the dynamic range of images. The experimental results of network structures with or without bypass verify the superiority and effectiveness of the proposed method.

Based on the traditional line-structured light measurement technology, we proposed a line-structured light imaging method of rail profile based on polarization fusion. First, the polarization component images and total intensity images of rail laser section at multiple angles were obtained by using a polarization camera. Furthermore, combining the correlation and complementarity of light strip information in the polarization component images and total intensity images, we constructed an image fusion algorithm based on the evaluation of light strip reliability, which improved the image quality of rail laser section. In addition, the experimental demonstration using the polished rails indicated that in the fused images of polarization component images and total intensity images, there was no longer local overexposure of light strip and the quality of light strip was improved. In comparison with the traditional method, the maximum measurement error of rail profile in the overexposed area was reduced from 0.31 mm to 0.10 mm, effectively improving the accuracy of rail profile detection.

Parallel translation computed laminography (CT) has higher imaging resolution in the direction of scanning and lower spatial resolution in the direction perpendicular to the scanning direction. We propose an orthogonal translation CL (OTCL) method to perform line scanning and realize high resolution imaging for objects to be detected in two orthogonal directions. A geometric model of the OTCL is established, the scanning motion is analyzed, the experimental system is established, and simultaneous iterations reconstruction (SIRT) algorithm is used to reconstruct CL images. Simulation and experimental results show that the OTCL method is feasible and can be used in the nondestructive testing of flat plate parts.

Imaging through scattering media has important applications in many fields. However, there are few studies on imaging and tracking of three-dimensional (3D) moving objects through scattering media. In this paper, a method of single-shot videoing for 3D moving objects through scattering layers is proposed. The speckles emitted from two different positions of the scattering layer are uncorrelated and contain the object information under dual view angles. Detecting in the overlapping area of the two parts of the speckles can realize 3D imaging. By rotating the camera around the optical axis during the exposure process, a speckle image consists of a series of momentary speckles with different rotation angles can be detected. With the speckle rotation decorrelation property, these momentary speckles are uncorrelated with each other. Thus, a series of momentary speckles corresponding to different moments and different view angles are multiplexed in a speckle image. Using the cross-correlation deconvolution imaging method, the video information of the 3D moving objects can be reconstructed from this speckle image by rotating the point spread function detected in each single view.

In this paper, a hybrid use of response surface approximation methodology and direct optimization method is utilized to design the lightweight and support point layout of the primary mirror. Taking the ultra-low expansion primary mirror of a 2.5 m ground-based optical telescope as an example, the parameter sensitivity analysis of this method, the global optimization of multi-objective genetic algorithm based on Kriging response surface, the local optimization process based on the gradient algorithm of the mixed-integer sequential quadratic programming are studied, and the evaluation function is formulated using the compromise programming theory. The results of integrated optimization show that, compared with the solid mirror of the same size, the lightweight rate is 72.13% when the primary mirror adopts a partially open hexagonal hole sandwich structure on the back. The primary mirror adopts 54-point whiffletree passive support in the axial direction. The root mean square of the mirror surface deformation under the vertical optical axis and gravity load is 6.08 nm, and all indexes meet the design requirements.

A design method of a primary-mirror position control system based on motor-driven hydraulic support is proposed to meet the high precision requirements of the primary mirror position in mirror imaging of a 4 m large telescope. Firstly, the composition of the primary-mirror position control system is introduced, and the mathematical model of each structure is established. Furthermore, a primary-mirror position controller is designed based on the first-order dynamic sliding mode control and a linearly extended observer. Finally, the system is simulated. The results show that the maximum tracking error of each support area is less than 0.5 μm when the tilt axis moves at a uniform speed of 1 (°)·s -1. Besides, in the case of sinusoidal guidance of the tilt axis, the maximum tracking error is 1 μm, which is better than the tracking error (13 μm) of the traditional proportional-integral control. The results satisfy the design requirements of the proposed system in the 4 m large telescope, providing a reference for the design of a primary-mirror position control system in a large telescope.

In the field of spaceborne long-distance non-cooperative target ranging, usually the target is weak and the imaging frame rate is limited, and thus the target tracking ability is limited and influences the ranging distance. When the spaceborne long-distance non-cooperative target is tracked, the main source of tracking errors is body vibration. In order to enhance the active vibration suppression ability and improve the tracking accuracy of the system, adaptive control is used to realize aiming control of targets. Aiming at the problem of difficult data fusion among different frequency sensors, we use down-sampling to achieve adaptive fusion of low-frequency camera data and high-frequency inertial sensor data, and simultaneously confirm the effectiveness and correctness of adaptive filtering under down-sampling. At the level of algorithm implementation, an affine projection algorithm is selected to achieve data fusion considering the problems of slow convergence and poor steady-state accuracy of the classical adaptive algorithm. In the data fusion simulation, the micro-vibration model of the satellite is used to verify the convergence speed and the tracking accuracy of the fusion method. Finally, this method is applied for the actual tracking system and the experimental results show that the tracking accuracy is superior to 5 μrad, which meets the tracking accuracy requirement of long-distance non-cooperative target ranging.

To solve the problem that the contact measurement method is susceptible to the limitation of the sensor acquisition channel and the additional weight, a wind turbine blade crack diagnosis method based on three-dimensional (3D) vibration information fusion and convolutional neural network is proposed. First, based on the principle of binocular photogrammetry, a multi-channel sample construction method of 3D vibration information fusion is proposed. This method can integrate the motion information of multiple measurement points on the surface of the wind turbine blade, gain the acquired signal with more abundant features, and greatly decrease additional weight interference. Secondly, in order to obtain multi-level semantic information of cracks, a new multi-scale convolutional neural network is proposed. A type of 1.5 kW wind turbine blade was selected to carry out crack diagnosis experiments, and a database of samples of different crack states was established. The prediction accuracy reached 93.4%, which verified the effectiveness of the proposed method. Comparative analysis with the classic LeNet-5 and VGG-11 networks shows that the improved convolutional neural network has higher identify precision and faster convergence speed. Multi-channel signal samples can offer a better effect in wind turbine blade crack fault diagnosis application.

We develop a quantum vacuum measurement device based on a Fabry-Perot laser resonant cavity and use the beat frequency and the Pound Drever Hall (PDH) laser frequency-locking technology to accurately measure the change of resonant frequency in the cavity before and after inflation. The refractive index of dry Ar at the temperature of 299.1485 K and the vacuum degree of 10 2-10 5 Pa is measured. With the measurement result by the capacitance diaphragm vacuum meter, we get the relation coefficient of 2.50835×10 -9 between refractive index and vacuum degree. In addition, we compare and analyze the difference between the retrieved and measured vacuum degrees, and finally, we evaluate the laser frequency uncertainty as well as the refractive index and vacuum measurement uncertainty. The results show that the laser frequency uncertainty is 5×10 -12, the refractive index measurement uncertainty is 1.64×10 -8, and the vacuum measurement uncertainty is 2.5×10 -4. The developed quantum vacuum measurement device based on the Fabry-Perot laser resonator has been used to realize the quantum vacuum measurement, which provides some reference for the research of quantum vacuum measurement technologies in China.

The saturable absorption characteristics of Co 2+∶MgAl2O4 crystals in different cutting directions were experimentally studied. The anisotropic transmittance characteristics of [100] and [110]-cut Co 2+∶MgAl2O4 crystals were measured, and the measurement results showed that the crystal transmittance of [100] and [110] cutting directions changed periodically depend on the polarization direction of the incidence light. And, the output characteristics of two cutting directions of Co 2+∶MgAl2O4 passively Q-switched lasers in [100] and [110] are compared. Results show that the [110]-cut crystal has obvious advantages. The output energy is about 8% higher than that of the [100]-cut crystals, and the stability, beam quality, and extinction ratio are improved slightly.

A frequency stabilized laser capable of long-term autonomous operation is the basis for the continuous operation of an atomic fountain clock. In this paper, based on a 780 nm commercial external cavity diode laser used in the 87Rb fountain clock, we develop a long-term automatic frequency stabilization system by using an embedded system. The proposed system can automatically identify the target peak in the 87Rb saturated absorption spectrum, stabilize the laser frequency for a long time and quickly re-lock it after an accidental loss of lock. In addition, we develop a method for dynamically adjusting the operating point. The target operating point changes with time because the influences of temperature, humidity, device aging, and other factors during the long-term operation of the laser. Based on this method, the laser frequency stabilization system makes lock loss rarely occur, and easily re-locks even if lock loss happens. The proposed frequency stabilization system has been successfully applied to our transportable 87Rb fountain clock. The clock can operate normally soon after transport and the long-term stabilization time of laser frequency can reach more than one month. The long-term stability of laser frequency relative to the 87Rb saturated absorption spectrum is about 2.3×10 -13 and the long-term stability of the transportable 87Rb fountain clock before and after transport is always at the magnitude of 10 -16.

Highly dispersed silica (SiO2) sol was prepared in a basic catalyst system using tetraethoxysilane, ethanol, and ammonium hydroxide as the precursor, solvent, and catalyst, respectively. SiO2 antireflective films with fourth harmonic generation antireflection at 266 nm were prepared by dip coating each film in different concentrations of SiO2 sol, and the properties of the as-prepared film were determined. Results show that the fabricated film pulled up at a speed of 4.5 cm·min -1 from the highest concentration SiO2 sol exhibited the best performance. The film was characterized by a maximum transmittance of 99.307% at 266 nm, surface roughness of 1.339 nm, and excellent stability.

Based on the chirped soliton solutions of the nonlinear Schr?dinger equation with cubic-quintic variable coefficients under Raman effect, the spectral evolution characteristics of chirped one-soliton and two-soliton in a saturable inhomogeneous fiber system with a periodic dispersion distribution are studied. The results show that in both homogeneous and inhomogeneous fibers, the high order effects do not affect the soliton waveforms in time domain, but directly influence their spectra, because they are only related to the phase of the solitons. In the homogeneous fiber, the chirp parameters of solitons are reduced to zero and the high-order effects make the spectrum of unchirped one-soliton redshift and the accompanied sidelobe appear at the side of high frequency. Compared with the one-soliton, the two-soliton with parallel transmission or colliding evolution shows splitting in its spectrum, and the high-order effects enhance the spectral splitting of two-soliton. Compared with the properties of one-soliton and two-soliton in a homogeneous fiber, the periodical change of chirp parameters in a periodic dispersion-distributed fiber causes that the pulse width and spectrum of one-soliton are periodically compressed and expanded. In contrast, as for the two-soliton, its spectrum is significantly broadened during collision and the chirp makes the spectrum of two-soliton blueshift for two-soliton with quasi-parallel transmission or colliding evolution, while the high-order effects make the spectrum redshift. The presented results are of great significance to study the spectral characteristics of high power signals in ultrafast optical communication systems.

By replacing the single-photon detector in a HBT (Hanbury-Brown-Twiss) experiment by a pair of homodyne detectors, we constructe a modified-HBT interferometer. We introduce the principle and procedure for the measurement of the second-order correlation function of light field based on the modified-HBT interferometer, and maily discuss the substraction of dark noises in the process. Based on the theoretical analysis, we realize the measurement of the second-order correlation function of displaced squeezed light output from an optical parametric osciallator and the characterization of photon statistics features through the simultaneous record and statistics of the quadrature information collected by the above interferometer for two spatially separated mode of light field. For amplitude squeezed light, its second-order correlation functional value is less than 1 under a specific displacement, indicating photon anti-bunching. In contrast, for phase squeezed light, the second-order correlation functional value is always greater than 1, indicating photon bunching. The properties of photon statistics for displaced squeezed light are explained from a viewpoint of two-photon interference, which can provide a reference for understanding the photon statistics features of squeezed light.

Modulation transfer function (MTF) is one of the important parameters for the image quality evaluation of optical satellite cameras. On-orbit MTF estimation has a bearing on the application of remote sensing data and the future development of remote sensing cameras. For the photoelectric imaging system widely used for earth observation at present, according to the physical definition of MTF, the satellite cameras are subjected to on-orbit MTF estimation through subpixel position detection and parameterized model fitting, during which the special point, line, and plane targets for convenient mathematical description are taken as references, such as reflecting point source arrays, radial targets, and large-area knife-edge targets. The experimental results show that the point source method is the most rigorous estimation approach and can fully characterize the imaging capability of remote sensing cameras. In contrast, with the assist of large-area targets, the square wave method can directly obtain the MTF value at the Nyquist frequency of the imaging system, and the knife-edge method is a common estimation method of optical cameras but can only have access to the MTF in the orbit direction and that perpendicular to the orbit direction. On-orbit MTF values obtained by the above three methods are of good consistency and the maximum relative error is smaller than 6.00%. At the same time, the above methods have their own characteristics and varied applicability.

Existing artificial local measurement methods are often used to evaluate the degree of surface weathering of grottoes. However, such methods are inefficient and the evaluation results are easily affected by subjective factors. In this paper, an intelligent quantitative evaluation method for grotto surface weathering based on multispectral imaging and random forest algorithm was proposed. Multispectral imaging was used to extract the the surface spectral information of grotto to characterize the type and degree of weathering. The multispectral feature data were reorganized and normalized to establish training, testing, and prediction samples. Based on the theory of minimum relative entropy, a loss function was designed to train a random forest algorithm model, and the spectral characteristics of samples with different weathering types and degrees were extracted. The weathering degree of each pixel in multispectral images of grottoes was predicted and evaluated using a classification model with feature perception ability after training. The confounding matrix and Kappa coefficient were used to evaluate the accuracy of the results. The proposed method was verified taking the Wanfo temple grottoes, Qingliang mountain, Yan'an city, Shaanxi Province as an example. Results show that the target grottoes’ strong salting-out weathering surface area ratio was 5.15%, weak salting-out weathering area ratio was 27.88%, slight salting-out weathering area ratio was 27.39%, and strong dust weathering zone ratio was 39.58%. The evaluation results were basically in accord with actual weathering conditions. Accuracy was 98.49% and the Kappa coefficient was 0.98. The proposed method can realize pixel-level refined evaluation.

Functional near-infrared spectroscopy (fNIRS) has attracted widespread attention as an emerging neuroimaging technology. However, the existence of motion artifacts in the fNIRS signal leads to bias in its signal processing outcomes. We proposed a tMedMor algorithm that combines the targeted median filtering (tMed) and mathematical morphology (Mor) for the removal of three motion artifacts in the fNIRS signal, namely, spike, baseline shift, and slow drift. Simulated and experimental data were used for verification, and the performance of the proposed algorithm was compared with those of several other common algorithms. Our results revealed that the tMedMor algorithm demonstrates good performance in terms of mean square error, signal-to-noise ratio, square of Pearson correlation coefficient, and peak-to-peak error, which together indicate that tMedMor can be applied as a new approach to the fNIRS signal at the preprocessing stage.