Taking a dual-band spectrometer as an instance, a stray radiation suppression device was designed for a compound optical system which began with a Cassegrain structure. The device contains outer baffle 1,outer baffle 2 and two inner baffles, in which the baffle 2 efficiently blocks the external stray radiation directly into the optical system. The device can effectively suppress the stray radiation of each spectrum, and with this device, the difference between parasitic light and signal light in visible branch are 7~8 orders of magnitude, which in near infrared branch are 5~6 orders of magnitude. In practical, the device performs well as the image is not affected by the stray radiation, as well as, the working time can be enlarged.

In recent years, visual imaging technology has been developed rapidly, the imaging system is also improving. Single image sensor has been unable to satisfy the demand, which is the large field of view and high dynamic range. Thus, multi-camera systems with large field of view(FOV) have attracted wide attentions. Based on the principle of compound eye imaging, the multi-camera system was used to study its imaging characteristics. Considering the loss in the imaging process and the influence of the lens, the irradiance of the camera was calibrated and the vignetting was corrected. At the same time, the functions of the cameras radiation response and dark current were solved. Then the large FOV image mosaic and high dynamic range imaging of multi-camera system were carried out. Experiments show that radiometric calibration has a good influence on the above imaging.

Advancements in laser decoy techniques have rendered laser semi-active guided weapons using traditional encoding methods unable to fulfill the needs of modern warfare. A fusion guidance system based on multiple sub-beam interference is proposed. The spatial “watermark cipher code” is combined with the temporal coded signal to produce a new temporal-spatial watermarking coded signal. The laser is divided into several equal power sub-beams which can produce an effective spatial watermarking interference pattern due to their coherence. Calculation results show that the interference pattern presents special distribution features of alternating brightness and darkness. The operating range and the identification of the guidance system are enhanced by the watermark encoding mode. When the position offset and angle deviation of the beam splitter, in the multiple sub-beams coherent emission system are less than 0.1λ(λ is the wavelength) and 0.1θ0(θ0 is the beam divergence), respectively, the effect of mechanical vibration on the emission system can be ignored.

To improve the real-time performance,the accuracy and the robustness of the unmanned aerial vehicle (UAV) autonomous navigation system under time-varying noise circumstance, an integrated navigation system of machine vision and inertial guidance was established, and an adaptive square-root unscented Kalamn filter (ASRUKF) was proposed. By introducing the Mahalanobis distance of the residual between the observed and predicted values to the minimal skew sampling square-root Kalman filter, the new algorithm can restrain the system noise adaptively, compute faster and be more robust to noise. The simulation results show that compared with the SRUKF, the ASRUKF is more robust to noise, and the computation time is reduced by about 38.8%, the forecast accuracy of displacement, velocity and attitude angle increases by more than 4 times, 4times, 6 times, respectively.

Compared with the traditional intensity image, the polarization image contains more abundant information of the reflection and scattering of the object surface. The underwater polarizing image was acquired by SALSA camera under natural illumination. The influence of underwater objects, depth, milk concentration and detection band on polarization imaging detection was studied. The results show that in the blue band, the polarization imaging can obtain better information of the boundary contour of the underwater object. The objects of different materials present different polarization characteristics under the water,and the copper has the highest degree of polarization 0.69. Under the condition of milk turbidity 1.40 mg/L, the polarization image can still detect the underwater target by comparing the target object's degree of polarization(DOP) information, the DOP of porcelain only reduces by 0.31. In addition, The image obtained by polarizing imaging is clearer than the intensity image at the depth of about 40 cm underwater. For example, the polarization contrast of iron is 5.26% higher than that of intensity.

For the cooled 640×512 detector with staring focal plane array, a mid-wave infrared (MWIR) continuous zoom optical system with 30× zoom range was presented .The design idea of the continuous zoom system consisting of the zoom objective lens group and the relay lens group without the rear fixed group was introduced. The image quality of the system at three positions of short effective focal length(EFL), middle EFL and long EFL is given; moreover,the amount of defocus and the distortion of image quality reflected in the full focal length range was analyzed. The experiment results indicate that the optical system with 3.7 μm～4.8 μm working waveband can realize 18 mm ～540 mm continuous zoom,the defocus amount of the full focal length is less than the depth of focus, the maximum distortion of the long EFL area is close to 0, and that of the middle EFL is less than 3%. It has the advantages of large zoom range, simple structure, short zoom path, and smooth zoom locus,which can be used for infrared optical observation and tracking system.

In order to ensure the whole field of view parallax can be measured, the charged-coupled device (CCD) camera which replaces eye must be located at the observation position (eye point) because the machine vision is not as flexible as the human eye, when the parallax of helmet mounted display is measured automatically. A pattern search method was proposed to make CCD camera locate at the observation position automatically in the optical plane of the helmet mounted display (HMD). Thus, the parallax of the helmet mounted display in the whole field of view can be measured automatically. The principle of the automatic measurement system was discussed in detail, and the process of CCD camera aligning observation point automatically was introduced in detail. Then the measurement precision and efficiency were also analyzed experimentally, as well as the alignment accuracy and repetitive positioning accuracy. The experimental results show that the method can ensure the CCD locate at eye point accurately with high efficiency. The positioning accuracy is ±0.071° .Compared with the pendulum method to measure parallax, it has high measuring efficiency and high repetitive positioning precision.

Multidisciplinary and multi-objective optimization design method was applied in the researches of collimator mirror lightweight of wide field spectrometer of extremely large telescope. Lightweight design was operated independently on the basis of collaborative optimization which is one of the best feasible multidisciplinary design optimization approaches combined with multi-objective genetic algorithm to carry out the optimization, and the Pareto optimal set was obtained. The lightweight performance with different lightweight holes and the thermal stability of light-weighting collimator mirrors were comparatively studied. With triangle lightweight hole of collimation mirror, the light weighting rate is about 70% and the peak-valley (PV) value is 82.696 nm. With rectangle lightweight hole of collimation mirror, the light weighting rate is about 75.3% and the PV value is 107.03 nm. When the temperature of the collimator mirror changes about 10 K, the deformation of the collimator mirror is added doubled. The results indicate that the comprehensive evaluation of triangle lightweight hole is better than rectangle holes; the coupling and interdisciplinary relationships among the lightweight, the lightweight hole shape parameter and the deformation of mirror are comprehensively considered during the multi-objective optimization of collimator mirror lightweight structure. Thus the designers can choose the satisfactory optimization results according to their demands, so as to significantly reduce the product development cycle and costs.

A long focal length lightweight zoom optic system was designed with a focal length of 30 mm ~ 300 mm, an angle of view of 1.1 ° ~ 11.4 ° and an F number of 3.5. Due to the long focal length of the zoom system and the need to reduce the quality under the premise of control caliber, after the analysis of the zoom theory and the actual situation, the positive compensation was adopted, and the Zemax software was used to simultaneously optimize the image quality and design the lightweight zoom system. In the optimization process, the aspherical surface was added to simplify the structure and improve the image quality. Without affecting the illuminance of the image surface, the off-axis light was properly blocked, so that the effective aperture could be as small as possible, and at the same time part of the lens material was replaced, balancing the contradiction between high image quality and light weight, and finally the overall mass of the system reduced from 937 g to 584 g.Results show that the system quality is good, the on-axis modulation transfer function(MTF) is above 0.3 at 120 lp/mm , the off-axis MTF is greater than 0.2 at 120 lp/mm, the MTF for each field of view is greater than 0.5 at 40 lp/mm, and the distortion is less than 1%. According to the zoom motion equation, the programming calculation was done by Matlab software , and the cam curve reflecting the motion process of the zoom group and the compensation group was obtained. In the process of zooming, the image surface is relatively stable and the focusing is smooth.

The random scattering effect of scattering medium on light is an important factor that restricts its optical focusing and imaging. The optical phase conjugation technique can realize optical focusing and imaging through the scattering medium by conjugate reduction of scattered light field, where the acquisition of scattered light field phase is the core of conjugate reduction. We expounded the basic principle of polarization phase shift, combined the polarization phase shift with the phase conjugation technique, and designed a novel digital optical phase conjugation system based on polarization phase shift. A 633 nm HeNe laser was used as the system light source, and frosted glass was used as the scattering medium to conduct scattered light focusing experiments.The experimental results show that the device can successfully achieve optical focusing through the scattering medium, and the ratio of the focal point to the background light intensity can reach about 400 times.

The current polarization imaging method has the problems of poor real-time performance, low energy utilization rate and low resolution. We proposed a four-channel polarization imaging method of non-coaxial type. The method used four channels to acquire images of different polarization directions to fuse the degree of polarization image. After that, we simulated parameters such as channel spacing, channel parallelism and polarization detection angle error affecting the polarization imaging results of the system, and gave the parameter tolerance meeting the polarization imaging accuracy of 1%. Furthermore, we developed a four-channel polarization imaging system of non-coaxial type, and conducted the experiment for vehicles 50 m away outdoors. The experimental results verify that the system can effectively provide the polarization imaging for target.

The traditional error analysis of monocular vision pose measurement only considers a single error factor, which causing the problem that there is a large difference from the actual project. An error analysis method considering multiple error factors was proposed. Firstly, a pose estimation model for monocular vision was established based on the measurement range and camera parameter. Secondly, the camera internal error, lens distortion error, image points detection error and target error were introduced into the measurement model simultaneously based on actual error levels. Finally, the influence of suppressing parameter errors, improving camera resolution on the pose measurement results were obtained, and the most effective precision optimization method was found. In addition,a monocular vision pose measurement system for small manipulator hand-eye location was adopted for experiment. The analysis results indicate that reducing the radial distortion error and improving the camera resolution can effectively improve both the position and attitude measurement accuracy of this system. The pose accuracy is increased by 6.57%, 4.21% and 5.88%, 5.54% ,respectively.

For high reflective surface and small fillet structural workpiece, traditional optical measurement methods and equipment are difficult to meet engineering requirements. A spherical center offset measurement method was proposed based on the blue light scanner.We made sand-treated steel balls adhere to the workpiece and scanned the steel balls by using blue light scanner.Then we fitted the sphere center of the point cloud and offset the sphere radius along the normal direction of the workpiece surface to obtain the actual workpiece point.The standard cylinders fitting-diameter-deviation is 0.000 2 mm and the measurement accuracy is less than 0.03 mm by using the method. The result proves the feasibility of the method.Taking a section of polishing blade which fillet radius of inlet and exhaust edge is less than 0.1 mm as an example, the measurement accuracy is within 0.03 mm.This method is not affected by the workpiece material and it has great significance for the measurement of the workpiece which has high reflective surface and small fillet feature.

As a kind of spectroscopic instrument, the monochromator has important applications in radiation calibration of sensors. In the actual application process, the setting of wavelength and bandwidth has an important influence on the precise calibration of the sensor.The low-pressure mercury lamp is used as the wavelength standard light source. Through the method of characteristic line scanning, it is studied that in the process of sensor calibration, it has an important influence on the precise calibration of the sensor under different slit widths of the monochromator. The results show that when the exit pupil is the same and the width is changed at the same time, the maximum deviation of the wavelength is 0.17 nm compared with the typical calibration state (the exit slit width is set to 0.5 mm); when the incident slit is inconsistent with the exit slit, the maximum wavelength deviation is 0.18 nm compared to the typical calibration state; when the light source is not filled with the entrance slit and the full aperture is filled, the maximum error is 0.02 nm, and the effect is almost negligible. In the related experimental study of sensor spectral radiometric calibration, the calibration accuracy and accuracy evaluation of monochromator have important applications.

The high accuracy of Fabry-Perot (F-P) etalon interval d has a very important influence on the interference measurement. Based on F-P interference imaging information, combined with the local subdivision principle of the peak coordinate and the regression of circular equation, in particular, a new method of virtual plane pixel subdivision and signal smoothing was applied, and the concentric ring diameter Di was obtained accurately. Then the accurate calculation of the decimal part and the integer part of the interference order was realized, and the measurement of the interval d by excess fraction method with three-standard wavelengths was completed. In addition, the influence of subdivision and non-subdivision on measurement results was analyzed. With the virtual pixel subdivision, the interval d is (2 009.961 91±0.000 06) μm, and (Uε/k0) is 9.8×10-7.Experiments verify the effectiveness of pixel subdivision method, providing effective ways to improve the accuracy of interval d.

A high precision and full autocontrol real-time measurement and remote control covering visible near infrared multi-wavelength sun photometer PSR-2 was developed. It can realize real-time measurement and display of direct radiation irradiance of the sun, sky radiance (main plane, equal zenith angle), aerosol optical thickness, atmospheric column water vapor content and ozone content.And it has the functions of simultaneous detection, accurate temperature control, accurate sun tracking and other detection functions. PSR-2 was improved on the basis of PSR-1, which had been tested for long time desert wind sand and rainwater erosion. It is characterized by long time effective stable observation, more miniaturization, more convenient data processing ,and higher performance and price ratio. In addition,the Langley method calibration and temperature control performance test were carried out on the instrument in Dunhuang. Results show that the PSR-2 Langley calibration fitting linear correlation is higher than 99%,the constant temperature warehouse temperature is stabilized within (25±0.3)℃, the PSR-2 aerosol optical thickness and the large column water vapor content deviation are within 0.02 and 0.1, respectively,compared with the actual measurement results in Hefei and Dunhuang by the foreign industry standard of CE318. Finally the error analysis was carried on.

The illumination system with 3-primary-color light emitting diodes (LEDs) and light rod has the advantages of wide color gamut and small volume. However, there is a problem of low brightness when combining the 3-primary-color LEDs with the light rod. Firstly, the LED light source was selected based on the etendue theory, and then a set of free-form collimating lenses and collecting lenses were designed. The simulation results show that the light emitted by the collimator lens is concentrated within ±7.5°, and the energy utilization rate within the ±30°at the exit of the light rod is about 51.7%. Moreover,the designed lenses were manufactured and tested. The practical results show that the projection illuminance of the designed lenses set without the anti-reflection film can be improved by 1.6% than the original system with the anti-reflection film. It is expected that a 9.8% improvement of the projection illuminance can be achieved after the designed lenses are coated with the anti-reflection film.

Based on the principle of synchronous radiation focusing mirror, a bending mechanism was designed,the design theory of the bending mechanism was introduced, and the calculation process of the theoretical shape of the bending mirror was deduced.Meanwhile ,according to the design theory, the model of the bending mechanism based on flexible hinge was established, and its parameters were calculated, the effective length was 84.9 mm and the maximum torque was 259.8×10-3 N·m.Furthermore,the design and check of the bending mechanism were completed,showing that the material mechanics requirements of the bending mechanism could be meet. Finally, according to the error of the theoretical surface shape error, how different factors affect the variation of the surface error was analyzed by comparing the theoretical value of the bent mirror shape with the finite element simulation value.

In view of optical alignment of slit-lamp rotating drum, a new digital optical calibration method was presented. This method adopts the principle of transmission aligning system, adds the cross lines before and after the detecting optical path, judges the aligning quality by observing the imaging of two cross lines in the charge-coupled device (CCD) image and achieves the digital information through image processing at last. By extracting the midpoint coordinates of two cross lines, it can be judged whether the optical axis of the rotating drum optical system is consistent; by extracting the clarity of CCD image,it can be judged whether the rotating drum optical system is focusing . This method was used to align the rotating drum. The results show that the proposed method is a digital method to transform aligning information to digital information by using image processing technology, and its alignment error is less than 0.1 mm by digital image analysis.

In order to solve the problem of poor efficiency and repeatability during adjusting the attitude of spherical components in laser differential confocal components parameters measurement system, we designed a laser confocal automatic adjustment system for attitude of spherical component. Based on the laser confocal principle, we set up the relationship between the component misalignment and the adjusting value of electric four-dimensional adjustment mechanism, analyzed the attitude information of the component according to the airy spot position constantly measured by CCD detectors, and realized the attitude automatic adjustment by using the closed-loop feedback control algorithm. Finally, we built an automatic adjustment experiment device. The experimental results show that the shift adjustment resolution of the electric adjustment mechanism reaches 0.5 μm, and the tilt resolution reaches 10″. The adjustment system can quickly and stably adjust the attitude misalignment of the device under test to the error range, effectively improving the repeatability and efficiency of the adjustment. It is especially important for the laser differential confocal measurement system to realize the automatic measurement of spherical component parameters.

Polarizers are important components for many optical systems. Sub-wavelength dielectric gratings can be used as polarizers under normal incidence, and these have broad application prospects in high-energy laser systems. In order to investigate the induced damage characteristics of nanosecond pulse laser with wavelength of 1 064 nm for subwavelength all dielectric gratings, the particle swarm optimization algorithm was used to design the geometric parameters of the grating with rigorous coupled wave analysis. The calculation results show that the subwavelength grating polarizer has 0.5 nm bandwidth and a theoretical extinction ratio of 1 500 at wavelength of 1 064 nm. This grating was fabricated by ultraviolet(UV) exposure and ion beam etching, and its nanosecond pulsed laser damage threshold was tested. The test results show that the laser induced damage threshold of S polarized light is about 5 times of that of P polarized light, and all of them are greater than 5 J/cm2. The results indicate that the sub-wavelength all dielectric grating polarizers can be widely used in laser system under normal incidence.

Optical antireflection film is an important part of laser systems, and it is also most likely to be destroyed under laser irradiation. How to improve the laser induced damage threshold of antireflection film is one of the research hotspots. The relationship between different gradient antireflection films and laser induced damage threshold were studied without changing the target transmission and the total optical thickness of the films . Firstly, a hybrid inhomogeneous film was designed by MATLAB with hybrid design method, namely G/H1→H/L/A. Secondly, the inhomogeneous film was graded into different layers by the inhomogeneous film grading equivalent method. And using the plasma enhanced chemical vapor deposition (PECVD) technology, different inhomogeneous antireflection film structures (multilayer graded-index antireflection film structures and corresponding inhomogeneous film structures) satisfying optical performance indicators were deposited on K9 glass. Finally, the laser induced damage threshold (LIDT) of the films were measured. The results show that, without changing the target transmission and the total optical thickness of the films, the inhomogeneous antireflection film has a significant improvement in the LIDT compared to the conventional antireflection film. When the number of gradient layers increases, the LIDT of the inhomogeneous film structures decreases. And for the inhomogeneous films of the same layer, the LIDT of the sample prepared by the slope method is better than that of the sample prepared by the gradient.

It is able to generate the ultrasonic waves on the surface of materials based on the laser-generated ultrasonic technology which is an important part of non-destructive testing for material defects. With the finite element analysis software of Abaqus, the defect model of the symmetric aluminum to be measured was established based on the laser ultrasonic thermo-elastic mechanism. Besides, the propagation characteristics of the laser-induced surface wave in the material and the interaction process between the laser-induced surface wave and the defects were simulated. Through the numerical simulation and theoretical analysis, the results show that the amplitude of the reflected wave increases with the depth of the defect and reaches steady state when the depth of the defect is up to a certain value. Instead, the width of the defect has a limited effect on the reflected wave. All of these conclusions can promote the development of the laser-generated ultrasonic technology for the detection and identification of the material defects.

The impact of atmospheric turbulence on free space optical communication systems is not negligible. To reduce the influences of atmospheric turbulence on free space systems, a free space optical communication system based on wide-spectrum partially coherent laser in 900 m real atmospheric channel was experimentally demonstrated. The system used a modulated wide-spectrum partial coherence pulsed laser which was filtered from a supercontinuum to communicate, and the supercontinuum was gained by pumping picosecond pulses into a high-nonlinear fiber. In the process of testing, a reference link was set to ensure consistency of the testing environment. The results show that compared to the narrow linewidth communication system,under mid-turbulence condition, the system scintillation index is 0.053 8 which has 23% improvement and the system lowest detection sensitivity is -23.35 dBm which has 42% improvement .The wide-spectrum partially coherent optic communication system can significantly reduce the optical power jitter caused by turbulence and improve the communication performance of free-space optical communication systems.

A 5 Mbit/s 2-pulse position modulation(PPM) modulated visible light image transmission system was realized, using field-programmable gate array (FPGA) as the main control chip. The whole system was divided into two modules: transmitter and receiver, all of which were implemented by FPGA.Firstly, the processed picture signal was sent to the FPGA through the serial port by the host computer, and modulated into 2-PPM signal, the signal transmission data frame format was designed, then modulated onto the red light emitting diode (LED) and sent out. Secondly,at the receiving end, the synchronization and demodulation of the data signal were completed according to the data frame format, and the image recovered after demodulation was displayed on the liquid crystal screen. Experiments show that the system can realize the transmission of visible light image signals.

The measurement and control of dust concentration in mine environment is always a difficult problem for the safety development of mining field.The uniform spatial distribution of experimental dust was observed by an internal-cavity circulation system, and a two-optical fiber differential measurement method was adopted to solve the problem of dust precipitation on the surface of the detection window, both of which could contribute to the safe and real-time monitoring of dust concentration in the downhole location. Thus, taking cornmeal powder as the research example, the measured value of dust mass concentration was calibrated by calculating the ratio of the dust mass concentration to attenuation coefficient under 5 different injection masses, which led to a basically agreement of the measured results and the ideal calculated value in the curve. It is found that the two-optical fiber differential measurement method can effectively solve the dust problem on the surface of the detector and can provide a reference for the engineering application of dust-concentration optic fiber monitoring technology in the mine environment.

A detection and counting unit with optical fiber was designed for realizing the detection and counting of micro-droplets on the microfluidic chip. TracePro was used to simulate and provide the basis for detection signal processing. The system can realize the counting function according to the change of the optical intensity, which is caused by droplet passing. Simulation were carried out to analyze the influence of beam collimation, droplet size, droplet refractive index and distance from receiving fiber to chip on optical intensity. The results obtained from the simulation show that the poorer the collimation beam, the larger the droplet radius, the smaller the refractive index, and the closer the distance between receiving fiber and chips, which can cause more apparent variation in optical intensity. What’s more, the droplet size determines whether there are double valleys in the optical intensity curve. When the radius of droplet is less than 13 μm, there is one valley in the optical intensity curve. And when the radius is larger than 17 μm, there are two valleys.In addition, processing method of detecting signal was proposed based on the simulation results, showing that this optical detection unit can realize the detection and counting function of the micro-droplets on the microfluidic chip.