The theoretical model evaluation method of static target discrimination performance based on TTP criterion has advantages of strong environmental adaptability, accurate performance prediction and convenient system parameter optimization, but the accuracy of this method in evaluation of moving target detection performance needs to be further studied. In order to accurately evaluate the detection performance of moving targets, the target motion system, infrared target generation system, target acquisition and storage system were established. The experimental method of human eye threshold judgment was used to evaluate target detection performance at different imaging distances, different motion velocities, and different target sizes, and the target detection probability curve was obtained. According to the NVThermIP theoretical model for field performance evaluation, the target detection performance was calculated when the input of system performance parameters was consistent. The experimental results show that the NVThermIP evaluation model has certain deviation for detection performance prediction of moving targets. In the range of target motion rates measured by experiment, the target is easier to detect with the increase of the target motion rate. Research results provide data support for the revision of theoretical model of performance evaluation of moving targets.

To realize single-station passive location to sea surface ship target by photoelectric reconnaissance equipment, the method of high-precision and near-real-time photoelectric single-station passive location to sea surface target was proposed. The equations of single-station passive location were constructed, the location errors were analyzed, the factors affecting the location errors were explored, and the application suggestions of single-station passive location were given. Simulation results show that the location error is about 1.672%R at range of 17.899 km, with once location time about 20 ms. The errors of location, latitude, longitude and altitude are separately subject to normal distribution with N (299.31 m, 197.39 m), N (4.944×10?5°, 2.665×10?3°), N (4.038×10?5°, 1.899×10?3°)and N (?0.485 m, 14.165 m). The errors of pitch, azimuth and target altitude have a great influence on the location accuracy. The location results are reliable when pitch is in the domain of [?120°, ?15°], and the location error is smaller than 800 m when target altitude error is in the domain of [?200 m, 200 m], which realizes the high-precision and near-real-time single-station passive location to sea surface target. The proposed method is suitable for single-station passive location of sea surface, ground and air targets with reference plane of target altitude.

To meet the requirements of mid-wave infrared radiation measurement and calibration, the micro integrating sphere and mid-infrared detector were integrated in cryogenic environment. Several samples used to make the inner cavity surface of the integrating sphere were verified and compared. Through the measurement of broad-spectrum reflectivity, characteristic spectrum reflectivity, bidirectional reflectance distribution function (BRDF), laser confocal microscope, scanning electron microscope (SEM) and other micro-surface morphologies, the inner cavity surface fabrication process suitable for making the mid-infrared integrating sphere was obtained. The BRDF test results show that it is close to the Lambert reflection surface. The spectral and electrical properties of the integrating sphere detector sample were tested. The spectral band of the sample after surface roughening treatment and evaporation of metal reflection film has good adaptability, the light intensity attenuation ratio is 0.067 4, and the reflectivity of the inner cavity wall of integrating sphere is 96.4% by calculation. At the same time, the introduction of the cryogenic integrating sphere reduces the noise of the detector chip from 3.5×10?12 A·Hz?1/2 to 1.0×10?12 A·Hz?1/2.

With the development of aerospace imaging technology, the demand for remote sensing imaging observation system is increasing. In order to simplify the alignment and assembly complexity of the optical system and reduce the volume of the optical system, a designed method for low-cost and miniaturized remote sensing imaging optical system was proposed. On the basis of all spherical lenses and low dispersion materials such as calcium fluoride, an optical imaging system with a focal length of 2 000 mm and suitable for 5.08 cm (2 inch) detectors was designed. The modulation transfer function (MTF) was greater than 0.41 at 50 lp/mm, and the volume was close to the R-C system. The spherical primary and secondary mirrors of the system avoided the use of aspheric compensators during alignment and adjustment, which reducing the cost. The structure of a new type of louver baffle was designed to effectively avoids the problem of excessive volume of the conventional outer baffle structure. After the avoidance angle is greater than 12°, the point source transmittance (PST) value is less than 10-6, and the optical volume is only ?227×421 mm3, much smaller than the conventional outer baffle form. The system offers advantages of low cost, miniaturized structure, light weight, excellent image quality, and good stray light suppression, which can meet the requirements of aerospace imaging system in engineering applications.

The polarization regulation technology can effectively fulfill the requirements of rapid and precise identification and detection in the visible light range for polarization microscopes. The key lies in enhancing the polarization extinction ratio and achieving the optimal utilization of light. A composited multivariable nanowire gratings structure was designed based on the equivalent medium theory and the resonance enhancement mechanism of the resonator. The relationship between the structure parameters and polarization property of the structure was analyzed by numerical simulation using the finite-difference time-domain (FDTD) method, which could realize the polarization regulation function of wide angle incidence, high transmittance and high extinction ratio in the visible light band. Combined with the existing conditions of the laboratory, the nano-imprint technology was used for sample preparation and experimental verification. The experimental and simulation results show that the transmittance and extinction ratio of TM mode are higher than 60% and 35 dB respectively within the incident angle of ±35°, and the error range is about 4% at the vertical incidence of 532 nm laser source. Compared with polarizers in existing polarization microscopes, the designed structure of polarization regulation ensures high transmittance while increasing the extinction ratio by 1.5 times.

Two different frequencies (1 kHz and 15 kHz) of medium-wave infrared lasers were used to compare the interference effect at a distance (15 km) to the same target, the relationship between laser frequency and interference effect was analyzed, and the laser power density reaching the detector was calculated. The experimental results show that the interference effect of the medium-wave laser with high repetition frequency (15 kHz) is obviously better than that of the medium-wave laser with low repetition frequency (1 kHz) when the laser wavelength and average power are the same. The experimental results have certain guiding significance for the development of photoelectric countermeasure equipment.

With the development of micro-nano machining technology, the chiral metasurface has been proved to have great application potential in optics. Traditional metal chiral structures are difficult to obtain ideal circular dichroism due to their own ohmic loss, and three-dimensional chiral structures with good circular dichroism but complex structure are difficult to manufacture. The two-dimensional all-dielectric chiral structure of periodic double-Si bricks was adopted. By exciting the non-orthogonal in-plane magnetic dipole moment and electric dipole moment, the optical chiral response could be generated, and the ideal circular dichroism could be obtained. Through simulation optimization, the structure can achieve ultra-high circular dichroism of more than 0.94 at an operating wavelength near 1 550 nm, and can realize the conversion of incident left-circularly polarized light to right-circularly polarized light. This all-dielectric two-dimensional chiral metasurface has the advantages of high circular dichroism, simple structure and low preparation difficulty, and has a wide application prospect in the field of near infrared polarization imaging.

Low-altitude small unmanned aerial vehicle (UAV) frequently invade sensitive areas, which posing a serious threat to national and social security. Aiming at the problems such as high missed detection rate and insufficient detection accuracy for UAV target detection based on thermal imaging, the infrared detection of UAV-YOLO (IDOU-YOLO) algorithm model was proposed. A cross-scale fusion feature pyramid mechanism was constructed to fully explore the feature space information, focus on the cross-scale information fusion and the rich information representation ability of the model, and enhance the target recognition ability. At the same time, the bounding box loss function Scylla IoU (SIoU) was introduced to improve the detection accuracy and accelerate the convergence speed of the model in the training process. The experimental results show that the precision, recall, F1 score, mAP@0.5 and mAP@0.5:0.95 reach 99.2%, 96.3%, 97.7%, 98.4% and 70.2%, respectively, which indicates that the IDOU-YOLO model improves the detection and recognition ability of UAV targets in various scenarios, and can better meet the application requirements of anti-UAV systems.

In order to effectively solve the problem that small objects are difficult to recall and easy to miss detection, an improved YOLOv3 algorithm based on feature fusion and feature enhancement was proposed. To enhance the generalization performance of the model, the Mosaic and Mixup methods were combined for data enhancement during training. Firstly, in order to improve the low recall rate of small object detection, the original feature fusion network was extended to the shallower layer, and the bottom-up feature pyramid was added, so that the details and positioning information of the shallow feature layer could be transmitted to the deep layer. Secondly, a feature enhancement module was proposed to enlarge the receptive field, so that the shallow feature layer could obtain the rich deep semantic information and optimize the expression ability of the feature layer. Finally, the GIoU was used as a regression loss function to reduce the missing rate and achieve more accurate regression. Simulation experiments on Pascal VOC2007 and VOC2012 show that the improved algorithm can improve the mean average precision (mAP) value by 4.4% on the premise of ensuring the detection speed. Experimental results fully prove that the proposed algorithm can effectively improve the performance of small object detection.

Semantic segmentation is a computer vision technology widely used in scenarios such as unmanned driving and defect detection, but the fine-grained annotation at the pixel level requires a huge annotation cost. Therefore, how to use the easily obtained image-level labels for weakly supervised semantic segmentation is the focus of long-standing research. Compared with pixel-level segmentation based on a class activation maps (CAM), a masked consistency mechanism (MCM) was proposed to provide additional supervision signals to narrow the gap between full supervision and weakly supervision. In the fully supervised semantic segmentation, the network had consistent pixel-level segmentation supervision for mask prediction of each patch of the image, so some patches were masked out in vision transformer (ViT) and it was required that the CAMs generated by the retained patches should be consistent with the CAMs generated by the complete images to provide additional self-supervision signals for network training. Experiments on PASCAL VOC 2012 and MS COCO show that the proposed method is superior to the most advanced method using the same level of supervision.

To solve the problems of aberrations and motion blur encountered in infrared imaging, an algorithm for infrared imaging deblurring based on spectrum analysis was proposed. By analyzing the blur problems, a blur function estimation approach was constructed by using the spectral information of infrared images, and the deblurring of infrared images was achieved by optimizing the data fitting terms and regularization terms. The experiments were conducted on datasets containing motion-blurred images and static aberration images. The results show that in the experiments of datasets deblurring, in comparison to previous methods, the peak signal-to-noise ratio (PSNR) of proposed algorithm is improved by 2.01 dB, and the structural similarity index (SSIM) is improved by 0.06. The static aberration images are acquired using a short-wave infrared detector under laboratory conditions, and the restored images show the significant improvement subjectively, particularly in high-throughput optical system imaging scenarios.

An improved U2-Net segmentation model was proposed to address the issues of weak feature extraction ability and low segmentation accuracy in metal coating defect image segmentation. Firstly, an improved receptive field block light (RFB~~l) module was embedded in the U-shaped residual block (RSU) to form a new feature extraction layer, which enhanced the ability to learn detailed features and solved the problem of low segmentation accuracy caused by limited receptive field in the network. Secondly, in the decoding stage of the U2-Net segmentation model, an effective contour enhanced attention (CEA) mechanism was introduced to suppress redundant features in the network, obtain feature attention maps with detailed position information, enhance the difference between boundary and background information, and achieve the more accurate segmentation results. The experimental results show that the average intersection and union ratio, accuracy, precision, recall, and F1-measure of the model on two metal coating peeling and corrosion datasets are 80.36%, 96.29%, 87.43%, 84.61%, and 86.00%, respectively. Compared with commonly used SegNet, U-Net, and U2-Net segmentation networks, the performance of the model is significantly improved.

Turbidity is of great significance for evaluating water quality. Addressing the issue of large measurement errors and low resolution encountered with the 90° scattering method in high turbidity scenarios, an advanced turbidity measurement technique utilizing 135° scattering was introduced, suitable for a wide range of turbidity. Since the scattering phase function was closely related to the scattering angle, the turbidity measurement adaptability of the scattering angle was studied by experiment. The analysis, conducted through polynomial fitting order, revealed that the 135° scattering method demonstrated a more extensive measurement range, particularly effective in the 100 NTU~2 000 NTU range. The method underwent rigorous experimental calibration, yielding segmented linear fitting R2 results of 0.99 for 100 NTU~800 NTU, 0.98 for 800 NTU~1 500 NTU, and 0.99 for 1 500 NTU~2 000 NTU. Experimental results show that the 135° scattering approach can extend the turbidity measurement range and improve the precision of measurements in high-turbidity conditions.

To meet the requirements of rotor axial center trajectory measurement in helicopter flight state, an axial center trajectory measurement method based on low-altitude binocular vision was proposed. Firstly, for the small-value measurement of high-speed rotating objects, based on the principle of stereo vision measurement, a visual measurement system of rotor axial center trajectory with unmanned aerial vehicle (UAV) as a low-altitude mobile platform was designed and constructed. Secondly, according to the environmental conditions of dynamic measurement, a measurement system calibration method combining ground calibration and aerial dynamic real-time calibration was proposed. Then, based on the spatial geometric relationship, the static estimation and dynamic trajectory measurement methods of rotor axial center were discussed. Finally, the ground accuracy verification test of the measurement system and the rotor axial center trajectory measurement test of the helicopter ground start-up state were carried out. Experimental results show that the method is accurate and reliable, and the distance accuracy between points is better than 1‰, which can realize the non-contact accurate measurement of the axial center trajectory, and can also be applied to other dynamic small-value measurement tasks.

A surface defect detection method for curved optical lenses based on improved YOLOv5s was proposed to address the issue of low accuracy in surface defect detection. Firstly, a device for collecting surface defects on lenses was designed, and the collected defect images were enhanced to create a dataset of lens defects. Secondly, in order to enhance the utilization of channel information in YOLOv5s network, the SE attention mechanism was introduced into the feature extraction network to enable the network to extract information more accurately, and Transformer was integrated into the last C3 module of the backbone network to help the network better extract global information and improve detection efficiency. Finally, considering the issue of easy loss of feature information for small targets, the feature layer of the backbone network 160×160 was added to the feature fusion of the neck to increase the utilization of shallow information by the network. The mean average precision (mAP) and recall (R) of the improved YOLOv5s object detection algorithm are 93.9% and 91.6%, respectively, which are 3.2% and 3.4% higher than that of the original network algorithm, indicating that the improved YOLOv5s algorithm can effectively detect surface defects on lenses.

The influence of refractivity wet term on atmospheric refraction effect is often ignored during the measurement flight data processing of partial cinetheodolites. Based on the 626 sets of radiosonde data in a certain area around the Bohai Sea in 2019, the deviation of height measurement by theodolite caused by atmospheric refraction was calculated with and without the consideration of refractivity wet term by using the ray-tracing algorithm. According to the statistics of the influence of refractivity wet term on the accuracy of height measurement at different elevation angles and different heights throughout the year, the results show that the refractivity wet term is an important source of height measurement residual errors. For the test tasks with high precision requirements for target height measurement, especially in the case of low elevation angle measurement, the influence of the refractivity wet term cannot be ignored.

Hardness is a performance index to measure the degree of softness and hardness of materials. In view of the characteristics of lutetium-aluminum garnet (LuAG), lutetium-gallium garnet (LuGG) and other optical crystals with small elastic compression, easy to break and unclear indentation boundary. An optical crystal Vickers hardness measurement method based on image processing was proposed. A deep learning YOLOv5s network was used to segment the image. The exact positions of the four vertices were obtained by adaptive binarization, maximum connected domain selection, skeleton extraction and then specific directional line segments detected by probabilistic Hough line treatment. The experimental results show that the average relative error can be controlled within 1.5%, which effectively reduces the calculation error of traditional network algorithm and traditional image processing algorithm, and is suitable for automatic and accurate measurement of optical crystal Vickers hardness.

The detection of electricity meter data in the distribution box of a power system provides important data support for power management and safe operation. Traditional manual methods for reading electricity meter data are inefficient and prone to errors, while existing deep learning methods are limited in model application due to large model parameter sizes. To address these issues, a lightweight and robust real-time electricity meter detection method was proposed. The parameter size of the model was reduced by reducing the number of layers and channels in the feature extraction network, and the lightweight of the network was achieved. While reducing the network parameter size, the global context and local multi-scale context were introduced to ensure the network feature representation and fitting capabilities. The global context focused on the position of electricity meter data in the meter box, while the local multi-scale context adapted to different sizes of meter data. Experimental results show that the proposed network achieves higher accuracy and faster detection speed than other detection methods, even with smaller parameter sizes.

Aiming at the difficulties in monitoring the galloping period of optical power grounded waveguide in the external field environment, a low-noise galloping monitoring array based on the input of reference sensor for weak fiber Bragg grating was proposed. The signal models for reference sensor and monitoring sensor were analyzed, and then the signal processing steps of adaptive filter through theoretical model were derived. Besides, by changing noise input with different correlation coefficients, the corresponding simulation results were calculated, so that the noise reduction effect of the monitoring sensor was compared and analyzed. Finally, the low-noise distributed galloping monitoring system based on weak fiber Bragg grating was deployed in a western autonomous prefecture with 220 kV power transmission line. The results of the field experiment show that, after self-adaptive filtering for the input signal constructed by the reference sensor, the average noise power spectral density is reduced by 6.34 dB within 100 Hz bandwidth, the monitoring intensity spectrum is optimized by the algorithm, and the galloping period events of about 100 m can be obviously observed within the monitoring interval. The research results create certain conditions for further improving the typical event classification data processing and pattern recognition database, and prove the effective prospects in engineering application.

In recent years, low-level-light image enhancement technology has attracted much attention, but there are still some problems. For example, sometimes dark areas are not completely improved, and sometimes bright areas near the light source are overexposed. In response to the above issues, an image enhancement method based on Retinex model with dual enhancement of illumination and reflection components was proposed. Firstly, the original image was transformed from RGB space to HSV space, and the V-component was extracted for subsequent processing. Then, the V-component was filtered to obtain the illumination component of the image, and the reflection component of the image was obtained by decomposition according to Retinex theory. Next, the global adaptive brightness enhancement was applied to the illumination component, and the multi-scale detail enhancement was applied to the reflection component. Then, the enhanced illumination component and reflection component were reconstructed according to Retinex model to obtain the V-component reconstruction image, which was processed by nonlinear transformation and local contrast enhancement. Finally, it was converted back to RGB space to obtain the final enhanced image. The experimental results show that the evaluation values of peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) of this method are 17.741 and 0.765, respectively, which have better image quality and better enhancement effect than other methods.

The organic membrane is used as a temporary substrate for preparing the Al2O3 ion barrier film, and its compactness and thickness significantly affect the compactness, the electrical transmittance characteristics of Al2O3 ion barrier film and the electrical characteristics of ion barrier microchannel plate (MCP). The non-solvent component was changed from ultrapure water to NaCl solution, and the compactness and thickness of the organic membrane prepared by different NaCl solution concentrations were analyzed by metallographic microscope and step test analyzer. The compactness, the electrical transmittance characteristics and the electrical characteristics of ion barrier MCP prepared by the corresponding organic membrane were measured. The research results show that when the concentration of NaCl solution is 0.05 g·ml?1, the organic membrane has high compactness, and the Al2O3 ion barrier film is compact and has small gain loss of MCP, which is the optimal non-solvent component concentration.

Ultrafast lasers, with the characteristics of narrow pulse width, wide spectrum, and high peak power, have important applications in many fields such as industry, medicine, and research. The dispersion compensation was carried out in the Figure-9 laser cavity, and the laser could achieve both wide spectrum output and self-starting mode locking. When the pump power was 160 mW, the continuous light component of the spectrum disappeared. The spectral full width at half maximum (FWHM) of the output was 22.6 nm and 26.2 nm respectively, and the compressed pulse width reached 143 fs and 128 fs. Under free operation, the root mean square (RMS) of the average output laser power within 1 hour was 0.1%. The designed laser has the advantages of wide spectrum, high integration, and strong stability, which can meet the application requirements of ultrafast laser in fields such as micro-nano processing, bio-optics, and spectral detection.

TiO2 and SiO2 films with different porosity were prepared on quartz substrate and Si substrate respectively by sol-gel technique combined with spin coating method, and the optical and laser damage characteristics of both films were investigated. The optical band gap of the films was calculated based on the transmission spectral curves. It was found that the optical band gap increased with the increase of porosity, the optical band gap size of TiO2 films ranged from 3.75 eV to 3.97 eV, and that of SiO2 films ranged from 3.52 eV to 3.78 eV. The ellipsometry measurement results show that when the mass fraction of polyethylene glycol (PEG) was increased from 0, 0.8%, 4% to 8% at a wavelength of 1 064 nm, the porosity of TiO2 films increased from 11.5%, 14.1%, 30.9% to 38.7%, and the refractive index decreased from 2.063 5, 2.016 5, 1.748 1 to 1.640 9; the porosity of SiO2 films increased from 4.04%, 4.6%, 5.7% to 13.9%, and the refractive index decreased from 1.438 6, 1.435 8, 1.420 4 to 1.387 9. The extinction coefficients of all TiO2 and SiO2 film samples are better than 10?3 except for the TiO2 film with PEG mass fraction of 0.8%, which indicates that the absorption of the films is smaller. The laser induced damage threshold (LIDT) of the films is greatly influenced by the porosity, and the larger the porosity, the higher the LIDT of TiO2 films, with a value up to 16.7 J/cm2. However, the LIDT of SiO2 film is not improved compared with that without PEG, and when the PEG mass fraction increased from 0.8% to 8%, the LIDT of SiO2 film increased by 1.9 J/cm2. In summary, improving the porosity helps to improve the laser damage resistance of the film.

Augmented reality (AR) system is a display system that superimposes virtual images on the real environment. Since the optical waveguide can realize miniaturization and light weight, it is the core component of the AR display system. The diffraction efficiency of the optical waveguide coupling element greatly affects the imaging quality of the AR display system. In order to improve the diffraction efficiency of the coupling element in the AR display system, a waveguide structure with multi-layer superimposed gratings was proposed. Based on the vector diffraction theory, the grating was simulated and analyzed by using the rigorous coupled wave analysis method. Simulation results show that the average diffraction efficiency of a single-layer inclined grating reaches 60% at a field of view of 50°, and the average diffraction efficiency of a waveguide structure with three layers of superimposed gratings reaches more than 85%, which is 25% higher than that of a single-layer grating. The analysis results show that the waveguide structure with multi-layer superimposed gratings can achieve a larger field of view and diffraction efficiency, so that the AR display system has a clearer viewing field.

A common reference calibration method was proposed for the precision alignment of off-axis two-mirror optical systems with adjustable focus, which ensured that the guide rail and the optical axis of the primary and secondary mirrors were precisely aligned with the same reference, and realized the parallel movement axis of the guide rail and the optical axis of the primary and secondary mirrors. The consistency debugging of the optical axis was completed through two parts: initial adjustment and fine adjustment. The initial adjustment realized that the primary and secondary mirrors were basically located in the theoretical installation position, reducing the adjustment errors. Then, the computer-aided adjustment was performed to achieve precise adjustment of optical axis consistency of the primary and secondary mirrors, so that the system wave aberration met the requirements. The experimental results show that the wave aberration root mean square (RMS) of the infinitely distant system reaches (1/15) λ (λ=632.8 nm), the system has a diffraction resolution of 0.64", and the system is focused to 10 mm. The measured focal length of the primary and secondary mirror system is 1 608 mm, and the system resolution reaches 2.6", which meets the design requirements.

A unique double-cladding photonic crystal fiber structure was designed, consisting of two layers of circular air holes in the outer cladding and two layers of ortho-octagonal air holes in the inner cladding, and the core material was a multi-component glass with rare-earth doped bismuth. The finite element method combined with scattering boundary conditions was used to analyze the mode field area, limiting loss, effective refractive index, time-averaged power and other characteristics of this structure. The results show that the corresponding mode field area and limiting loss at wavelengths of 1 310 nm, 1 550 nm and 1 600 nm is 726.5 μm2, 1.24×10?8 dB.km?1; 731.26 μm2, 1.32×10?8 dB.km?1; 732.28 μm2, 1.72×10?8 dB.km?1, respectively. The time-averaged power z-component at wavelength of 1 550 nm is up to 313 W·m-2, which can be used as a reference for gain medium of high-power lasers and provide a new option for wide-band tunable lasers.

A photonic crystal fiber sensor based on surface plasmon resonance was proposed for real-time detection of weak magnetic field. The surface of the photonic crystal fiber was covered with silicon nitride and silver film to achieve surface plasmon resonance, and the magnetic fluid on the outside of the sensor was used to sense the magnetic field changes. For the detection of magnetic field intensity, the performance indexes of magnetic field and sensor were summarized based on the performance indexes of refractivity. The simulation results of finite element method show that the wavelength sensitivity of the sensor can reach 0.55 nm?Oe?1 in the magnetic field range of 80 Oe~1 000 Oe. The amplitude sensitivity, resolution, limit of detection, quality factor and fitting degree reach ?0.107 Oe?1, 1.8×10?2 Oe, 3.3×10?2 Oe2?nm?1, 6.73×10?2 Oe?1 and 0.991 17, respectively. The sensor also has the characteristics of simple production process and easy to manufacture, which can foresee its broad prospect in magnetic field detection.