A method of the hybrid speckle-pattern compressive computational ghost imaging scheme is proposed. The scheme detects the larger and smaller resolution areas of the object via identifying complex object composed of different resolution scales automatically. The hybrid speckle pattern composed of different sizes of speckles is generated according to the identify areas. The compressive sensing techniques are combined to process the reconstructed image. Theoretical analysis and numerical simulation show that compared with the traditional computational ghost imaging, this scheme overcomes the influence of inappropriate selection of speckle sizes on the quality of reconstructed image, enhances the image contrast-to-noise ratio and visibility significantly, mean square error effectively. The scheme improves the image quality while reducing the sampling time, which further facilitate the practical application of computational ghost imaging.

In order to realize the photoacoustic microscopic imaging of strong scattering sample, a new method for scattering photoacoustic imaging is presented. A scattering photoacoustic detector is successfully designed by applying the scattering photoacoustic effect principle. The scattering photoacoustic detector consists of a scattering photoacoustic cavity, a resonant cavity, a microchannel and a microphone. The detector is used to detect photoacoustic signals of samples. Then, combining confocal scanning imaging technology, the photoacoustic microscopy imaging of strong scattering samples are achieved. Photoacoustic microscopic images of various kinds of scattering samples, such as silica microspheres and oral epithelial cells are successfully obtained as well. The results indicate that scattering photoacoustic imaging technique can greatly improve the image contrast and enhance image edge. This technique has important application meanings of particle diameters measurement in the field of industry and atmosphere.

Zero-order drift is inherently present in the interferogram of a spatially-temporally modulated ultraviolet Fourier-transform imaging spectrometer. The calculation of optical path of beam splitter in interferometric structure indicates that the zero-order drift is caused by refractive index variation in the ultraviolet wavelength range of optical adhesive and prism in the beam splitter. However, assembly misalignment of the prism may cause more serious zero-order drift, resulting in loss of the effective interference information and distortion of the spectral curve. The numerical simulation results verify that it is necessary to improve the assembly precision of the beam splitter to avoid loss of effective interference information. The measurement results show that the spectral inversion curve of the spatially-temporally modulated interferometric imaging spectrometer which contains inherent zero-order drift is consistent with the measurement result of a high-resolution spectrometer. When the beam splitter has high assembly precision, the zero-order drift does not affect the measurement accuracy of the imaging spectrometer.

To solve the non-uniform background problem induced by the illumination with uneven light intensity in the system of detecting surface imperfection on large-aperture optical components based on the machine vision, an image background correction method is adopted based on the mathematical morphology. An image segmentation algorithm of image gradient in combination with Otsu method is proposed. The experimental results show that the proposed method has the good anti-noise performance and high extracting accuracy for the blurred images at a certain degree.

One novel variational Retinex algorithm for infrared image enhancement is proposed to eliminate the staircase effect and enhance details. Gaussian curvature regularization is adopted to construct the variational Retinex model, and the first-order differential term is adopted to add the detail enhancement constraint, which realizes the adaptive enhancement of detail information. In combination with the neighborhood difference operation, the curvature filter solver for variational model is introduced to achieve the optimal solution. The experimental results demonstrate that, compared with those processed by the other variational Retinex algorithms, the quantitative evaluation indexes of images processed by the proposed novel algorithm are superior. The new algorithm can eliminate the staircase effect and enhance the details and visual effect of images.

Aimed at the difficult problem of segmented telescope co-phasing, a method of co-phasing error measurement based on dispersed fringe sensing technology is proposed. According to the detecting principle of the dispersed fringe sensing technology, an optical imaging model of dispersed fringe sensor is introduced. The simulation study is carried out on the computer. An auxiliary scheme called dispersed Hartmann measurement is brought out to solve the problem that dispersed fringe sensing algorithm loses efficiency under the near co-phasing condition. Simulations are also made to verify the feasibility of this method. By combining the two methods, dispersed fringe sensor can detect piston error under the range of ±60 μm in the visible waveband, and the detecting accuracy can achieve λ/10. Besides, the impacts of various factors on the detection accuracy have been analyzed in quantity. And solutions such as the fringe windowing, multi-trace and enhancing wavelength calibration accuracy are put forward to improve the dispersed fringe sensing technology performance. An advanced algorithm is also presented to make the method less sensitive to calibration errors. The results show that the proposed method can effectively detect piston error with high precision, no blind area and extremely wide range. It can be widely applied in the coarse co-phasing calibration and phasing control for space and ground based segmented mirror telescopes.

Comparing with traditional transmission imaging lens, the Offner refractive imaging lens has the advantages of wide band, small aberration, relatively large numerical aperture, compact structure and better space adaptability. In combination with the characteristics and requirements of spatial heterodyne spectroscopy (SHS), the Offner imaging lens and the light path structure are designed. Several pieces of lens are added to the exit light in order to change the pantograph ratio, while the aberration caused by the lens is eliminated by using off-axis assembly, so the requirement of variability in the magnification of interferometric fringe on localization plane on the SHS can be satisfied. Under the precondition of SHS requirements of a 0.042 numerical aperture and a -0.86221 magnification, the design example is provided, and the biggest root mean square radius of spot diagram is smaller than 1.81 μm. Finally, the imaging quality of the system is evaluated and the tolerance is analyzed, the result shows that the imaging quality can meet the imaging lens requirements of SHS, the Offner imaging of SHS can be achieved in engineering under the level of conventional machining and assembly.

The backprojection filtering reconstruction algorithm is proposed based on M-line for the circle-plus-line trajectory. The concise and appropriate M-line selection method is proposed, combined with the geometrical characteristics of the trajectory. The backprojection integral interval is obtained through solving the R-line coordinates which pass through the reconstructed point on the M-line, and the proposed reconstruction formula is derived. It is proved that the R-line within the reconstruction region for the circle-plus-line trajectory is unique, and the shortest length of the needed line trajectory is solved. Experimental results show that the proposed method can improve reconstruction image quality in large cone-angle scanning.

The computed tomography setup for the opposite parallel linear scanning (OPLCT) of X-ray source and detectors is simple in structure and feasible to achieve portability or mobility. The order subset-simultaneous algebraic reconstruction algorithm has been used in the preliminary studies, but its image reconstruction time is too long to realize fast imaging. To address this problem, we proposed a filtered back-projection algorithm for OPLCT (OPLFBP) based on the Fourier’s theorem. A single translation (1T) model was constructed for image reconstruction, while this model results in limited angle problem and cannot completely reconstruct the object image using the OPLFBP algorithm. Further, a multiple translation (MT) model was proposed, and meanwhile, models for two orthogonal translations (2T) and three symmetrical translations (3T) were constructed respectively. The simulation results for the 1T, 2T and 3T modes show that the OPLFBP algorithm is effective, and the 2T and 3T scanning modes and the traditional circular scanning mode are comparable in terms of reconstrution time and image quality.

The wide range for power dependence of phase shift spectrum of stimulated Brillouin scattering (SBS) is characterized through mathematical modeling and simulation; a phase shift measurement system of SBS loss based on heterodyne pump-Stokes technique is designed. The phase shift spectrum of 400 m standard single-mode fiber in SBS loss is measured in a pump wave power range of 5 μW~15 mW and a Stokes wave power range of 3.5~110 mW. The mechanism of measured center-asymmetric phase shift spectrum in SBS loss is analyzed. The results indicate that phase shift range of SBS loss has a good linear relationship with Stokes wave power and is almost independent of pump wave power. The measured center-asymmetric phase shift spectrum of SBS loss is mainly caused by fiber dispersion and nonlinear effect induced by nonlinear refractive index. According to the theoretical and experimental results, the applications of power dependence of SBS phase shift spectrum in microwave photonic signal processing and distributed optic-fiber sensing are analyzed, which provides theoretical basis for the design of application systems based on SBS phase shift.

The uplink physical throughput potential of orthogonal frequency division multiplexing passive optical network (OFDM-PON) is discussed. Reflective semiconductor optical amplifiers (RSOAs) are used in optical network units (ONUs) as the colorless resources to produce the dual-sideband intensity modulation, and a PIN photodiode with semiconductor optical amplifier (SOA) as pre-amplifier is used in optical line terminal (OLT) to complete direct detection. Based on a coherent narrowband integral model of RSOA, the weak perturbation solutions of the carrier rate equation and the wave propagation equation are used to investigate the RSOA’s dynamic electro-optical modulation behavior and frequency chirp characteristic theoretically and experimentally. It is proved that the RSOA is of low-pass modulation characteristics, and its 3 dB cut-off frequency is related to the bias current and the inject optical power. The optical equilibrium is achieved and the 3 dB cut-off frequency of the uplink can be promoted from 1.5 GHz up to 2.2 GHz by taking advantage of RSOA’s optical frequency chirp and the high-pass optical gain characteristic of SOA, and cascading a blue-shifted optical band-pass filter with RSOA. Furthermore, by utilizing the adaptive bit loading modulation, the data throughput in one RSOA-based uplink can reach 4 Gb/s and above，which means that the total uplink throughput of an OFDM-PON with ten RSOA-based ONUs can achieve 40 Gb/s.

An all-fiber optical path difference adaptive control method is proposed, and it has been successfully applied to the femtosecond fiber laser coherent polarization combination system. Field programmable gate array (FPGA) control circuit is used as the hardware foundation in the establishment of the optical path difference adaptive control system, and an effective control algorithm is designed based on the hill-climbing algorithm, which can successfully solve the disturbance problem between optical path difference adaptive control and phase-locked control. The experimental results show that the system can completely compensate the optical path difference drift in the range of ±60λ with good continuous working stability, which can meet the requirements of optical path difference adaptive control in the femtosecond fiber laser coherent polarization combination system.

An optical fiber sensing technology based on loss vector Brillouin optical time domain analysis (VBOTDA) is proposed. By constructing a heterodyne detection based simulation experiment system, the amplitude loss spectrum and phase shift spectrum of stimulated Brillouin scattering (SBS) are measured, and the temperature measurement of 50 m long sensing fiber is realized. The results show that the Brillouin frequency shifts obtained by SBS phase shift spectrum measurement agree basically with that obtained by amplitude loss spectrum measurement, and that both show a good linear relationship with the temperature. The temperature coefficient of Brillouin frequency shift obtained by loss VBOTDA is 1.16 MHz/℃, which is well consistent with the result of 1.2 MHz/℃ obtained by conventional BOTDA system. According to the theoretical and experimental results, the advantages of the proposed technology are analyzed.

In astronomical fiber spectrograph, the output field of step-index cylindrical multimode fiber under certain circumstances can become a ring profile rather than a spot with Gauss distribution. The change can greatly affect the measurement of astronomical spectroscopy, deteriorate the precision of spectrograph, and make the astronomical spectrum too complicated to analyze because of pseudo drifts and structures in spectrum. The propagation characteristics of step-index cylindrical multimode fiber are analyzed by ray optics, and thus the two major factors causing the formation of the ring output field are revealed. One is the coupling deviation, including center deviation and angle deviation. The other is the trace of the ray due to the rotation invariance of cylindrical fiber. A mathematic model is established to study the dependence of the distribution of output field on the center deviation. The formation and evolution process of the ring output field as the center deviation increases are studied, and the results show that they are in accordance with experimental results.

Given the fact that traditional linear block code identification methods cannot be applied to low density parity check (LDPC) codes, an ant colony algorithm is adopted to optimize the dual space search, so as to realize the LDPC code identification. The atmospheric laser communication channel model and the LDPC codes identification model are established, and the logarithmic likelihood ratio function of calibration relationship under the turbulent atmosphere channel is given. Then the basic ant colony algorithm is combined with LDPC code identification, the logarithmic likelihood ratio function is transformed into the objective function, and the recognition of LDPC codes is realized through continuous iteration for optimal value and optimal search path in the process of ants searching. The simulation results show that under the condition of 256 code length and weak turbulence, when the signal-to-noise ratio (SNR) is not less than 8 dB, the recognition rate can reach 78%; under strong turbulence, when the SNR is not less than 10 dB, the recognition rate can reach 77%. In addition, the parameter settings in the ant colony algorithm have a great influence on the algorithm performance and should be chosen according to actual situations.

The thermal damage analysis model of a classical circular cone protuberance on the surface induced by laser is set up. The modulation light field distribution and the temperature field distribution around the defects are calculated by finite-difference method, and the influence rule of the defects on thermal damage threshold is obtained. Results show that the thermal damage threshold increases while incident wavelength of laser increases and the size of defects decreases. In conditions of the same wavelengths, thermal damage occurs first in the defect when there is a defect, while thermal damage occurs in the middle region of the two defects first and it is more likely to make damage when there are multiple defects. For different wavelengths, the laser wavelength of 800 nm is more likely to cause damage in the material than the laser with wavelength of 1064 nm.

With Q-switched frequency-doubled laser at the wavelength of 532 nm as light source, the experiment of high speed laser scribing on amorphous silicon solar cell panels is conducted. By atomic force microscope, the scribing kerf depth, heat-affected zone dimension, kerf angle, and line energy density are measured, and the influence of scribing result on the solar cell efficiency is analyzed. The results show that the scribing result is ideal at the scribing speed of 0.3 m/s.

High efficient hundred-picosecond laser pulses amplification can be realized by scheme of stimulated Brillouin scattering (SBS) energy transfer. The amplification scheme can be used in the acquisition of shock pulse in shock-ignition. By exciting acoustic field in the liquid media, the process of energy transferring from long pulse to short pulse is realized. A SBS coupled wave equation under the transient condition is used to simulate the amplification process of 200 ps pulse in theory. The results show that SBS energy transfer scheme can amplify 200 ps laser pulse to a higher power density, and the amplification process accompanies with a certain pulse width compression.

The theoretical and experimental research on high power slab laser with dual doped concentration are conducted. By means of segmented doping, the non-uniformity in absorption pumping power density along the length of laser slab is effectively reduced, the average energy-storage density of single laser slab is obviously enhanced, and the total energy-storage increases by 39%. When the total pumping power of the diode is 15 kW, after passing through the dual doped laser slab, the 3 kW seed light source can extract a power of 5.16 kW, which increases by 36% if compared with that for single doped laser slab. The optical-to-optical conversion efficiency is 34.4%, which is basically consistent with the theoretical prediction.

Based on machine vision, one method used for detecting the surface defects of ceramic bowls is proposed. This method is mainly used to accomplish the edge detection of surface defects by means of Kirsch operator in combination with Canny operator. It adopts the eight direction templates from the traditional Kirsch operator to calculate the derivative convolution of every pixel point, singles out the optimal template, and determines the edge direction. The high signal-to-noise ratio, high detection accuracy and edge detail keeping of the Canny operator are combined to accomplish the detection of surface defects. Whether there is a defect or not is determined based on the the geometric features of defects. Experimental results show that this new method can effectively suppress noise disturbance, improve the edge localization and detection accuracy, keep the edge information but simultaneously avoid the occurrence of false edges.

To overcome the interference of the sky background in infrared image enhancement, and highlight the target in the image, an infrared image enhancement method based on pulse coupled neural network (PCNN) segmentation and fuzzy set theory is proposed. The PCNN is utilized to segment the image into sky background region and target region. The adaptive fuzzy enhancement method based on ridge type distribution is used to enhance the target region reflectance image obtained by variation Retinex, and the enhanced reflectance image and the illumination image are fused together. The local average value of target region is assigned to the sky background region, and the enhanced image is acquired by the reconstruction of the target region and the sky background region. Experimental results demonstrate that the proposed method can overcome the noise amplification problem of existing algorithm in sky region, and the enhanced images have high contrast and good visual effects.

A hybrid iterative algorithm which combines the transport of intensity equation (TIE) with the iterative angular spectrum propagation algorithm is studied. In the simulations and experiments, the phase retrieval performances by the TIE algorithm and the hybrid iterative algorithm at different defocusing distances are compared. The results indicate that, if compared with those by the TIE algorithm, the accuracy at large defocusing distance and spatial resolution of the phase retrieval by the hybrid iterative algorithm are improved. This algorithm provides a solution to the phase retrieval with high accuracy.

Aiming at improving the low accuracy of the binocular vision sensor along the optical axis in target tracking, the correcting method by one-dimensional laser range sensor is proposed. Meanwhile, in order to take full advantage of the binocular vision sensor and the laser range sensor in target tracking, the information fusion is used to improve the measurement accuracy and data utilization. In the process of measurement, the frequency of the laser range sensor is affected by the mechanical property of two-dimensional turntable, which causes the data of laser range sensor time-delay. Utilizing the direct updated algorithm to estimate the one-step prediction data and make it into real-time data in follow-up algorithm. Then using federated Kalman filter to solve the dependence of the multi-sensor data. Simulation results show that the real-time performance of the time-delay information is enhanced by the direct updated algorithm. Meanwhile, experimental results show that the final information fusion algorithm improves the accuracy of target tracking and improves the sensor shortages of the accuracy and timing.

The calibration method based on general imaging model is proposed by using the grating projection three-dimensional measuring technology for three-dimensional reconstruction of small objects. Different poses of the calibration target rotating around the fixed axis are utilized to calibration and optimization. The proposed calibration method solves the problems of general imaging model variables and does not need to use precision displacement device. In the meanwhile, a self-convolution blind deblurring method is used to solve the defocus caused by small depth of field of the telecentric lens, and improves the calibration accuracy of camera. The feasibility and veracity of the proposed method are demonstrated with experiments of calibration and three-dimensional reconstruction. The measurement accuracy achieves 6 μm when the camera field of view is 23.7 mm×17.78 mm.

The method of Raman-Nath symmetric beam splitting based on magnetically insensitive states and the scheme of internal state interference are proposed. The interference based on magnetically insensitive states is insensitive to the magnetic fluctuation, which is beneficial to the contrast enhancement of interference fringes. If such a Raman-Nath symmetric beam splitting method is applied in the path-conjugate atom gyroscope, the system noise is obviously reduced. By means of optimizing the symmetric-beam-splitting parameters of double pulses, it is possible to obtain a high diffraction efficiency. The scheme of internal state interference is successfully used to solve the problem of momentum state interference being susceptible to the environment.

In order to compare the influences on the observers color discrimination with normal color vision under different viewing conditions, the lighting sources with different spectral power distributions and displays with different primary colors are selected, and the four color matching experiments are carried out based on the monitor and printed color samples. The 47 color matching functions (CMFs) of Stiles-Burch are used for a computational color matching simulation and 27 to 36 observers with normal color vision are organized to carry out the color matching experiment. The mean color difference from the mean, observer metamerism magnitude and variability are used to quantify different observers color discrimination. The results show that the spectral power distribution of the monitor primary colors and the light sources have much effect on the observers color perception for color matching simulation, especially in gray color center, but it is not suitable for the color matching experiment. It indicates that the observers color discrimination is not only determined by the observer color matching function difference (observer metamerism), but also the visual mechanism and ability of color discrimination etc.

An atmosphere channel testing link is set up in the marine environment, and the experimental data are tested, including the atmosphere refractive configuration index, the variance of beam drifting and the beam radius. The comparison of free space optics bit error rate (BER) model with none of pointing error and the one with pointing error is carried on. Based on the BER model with pointing error, the BER performance of free space optics system is estimated with the refractive configuration index, beam drifting and beam spreading. Experimental results show that the actual measured BER increases with the atmosphere refractive configuration index. When the atmosphere refractive configuration index is small, the actual measured BER is larger 10 to 20 orders of magnitude than the theoretical calculated BER on account of the effect of beam drifting and beam spreading.

Whether correlators can output correlation peak is the main way to determine the target authenticity of correlators. However, it is prone to misjudgment when the accuracy of the components adjustment is not high enough. Based on the matched filtering correlators, we make theoretical analysis of three cases about the misjudgment of the correlation peaks. We give special emphasis to the theoretical and experimental study on correlation signal strength discrimination. Considering the three cases about the misjudgment of the correlation peaks, we explore the judgment of correlation peaks from the threshold intensity and the form of the correlation peaks, and the angle of rotation of the input image, which provide reference for experimental determination of the correlation peak.

The baffle size and Lagrange invariant of star sensor are two key factors which affect its volume and lens complexity. Through derivation and analysis of the relationship between limiting star magnitude and these two factors, miniaturization of the optical system of star sensor is studied. According to the relations between its optical system parameters and technological requirements, formulae about baffle size and Lagrange invariant under various limiting magnitudes are derived and analyzed. It is shown that with the increase of limiting magnitude, the Lagrange invariant monotonically decreases, but the baffle size has minimum value. On these grounds, limiting magnitude is optimized to get as small baffle size and Lagrange invariant as possible and to facilitate its miniaturization. As an example, the optical system of a star sensor based on STAR1000 complementary metal-oxide-semiconductor transistor (CMOS) detector and suitable for microsatellite is optimally designed, with limiting magnitude, field of view, entrance pupil diameter, optical overall length, and weight of 5.0 MV, 20°×20°, 28.2 mm, 124 mm, and about 300 g respectively.

In order to solve the difficulties of the multi-color, wide wavelength band, long working distance and high numerical aperture in imaging flow cytometer, an optical system of multispectral achromatic imaging flow cytometer is designed. Based on the modular design concept, the relative independence of each module is guaranteed by the flexible use of the local and whole optimization method. The microscope objective introduces diffractive optical element to correct chromatic aberration in the wide spectrum. So the imaging quality can meets ideal performance in full-wave band. The multispectral group uses dichroscope stack to split light, and six channels are optimized by multiple structure. The difficulty of chromatic aberration correction is reduced greatly. The final optical system is got by global optimization. The magnification is 60, the field of view is 60 μm×128 μm, the wavelength range is from 420 nm to 800 nm, the system has six channels, the resolution is 0.5 μm, and the imaging quality is close to diffraction limit.

Because of the flexibility of hollow waveguide, the waveguide with several meters long can be bent inside a round box with a radius of only several centimeters. By using light source and detector, the miniaturized gas sensing system with long optical path, small gas volume and fast response can be obtained. A curved waveguide transmission model based on geometrical optics theory is established. The transmission and sensing characteristics of bent hollow waveguide are investigated by simulation and experiment. The methods to reduce system detection limit and improve sensitivity by optimizing such parameters like the waveguide length, bending radius, system signal to noise ratio, waveguide aperture and light source divergence angle are summarized.

In order to reduce the weight and production cost of light emitting diode (LED) lamps, a special-shaped LED lamp without radiator is designed based on the principle of the chimney effect. A three-dimensional model is built by Solidworks software, and its plug-in called Flow Simulation is used to simulate the model. Based on the model that the height of chimney is 30 mm and diameter of chimney is 20 mm, the influences of different heights and diameters of chimney on the highest temperature of the special-shaped LED lamp are studied. Simulation results show that for the special-shaped LED lamp with the chimney height of 45 mm, chimeney diameter of 45 mm and substrate weight of 35.86 g. The highest temperatures of the LED are all lower than security junction temperature of 85 ℃, when the input powers are 6,8,10 W, which meets the security requirements. The experiment of special-shaped LED lamp with the power of 8 W is carried out. The result shows that the highest temperature of the special-shaped LED lamp is 73 ℃, which is only 2.06 ℃ higher than the simulation results. It confirms the correctness of the simulation. In conclusion, this special-shaped LED lamp without radiator not only has the ability to well satisfy the heat dissipation requirements of LED, but also has the light weight, a low cost and a simple manufacturing process.

For the problem of short communication distance in continuous-variable quantum key distribution (CVQKD) protocol and on the basis of multidimensional reconciliation scheme, the low-density parity-check (LDPC) codes with good degree distribution are designed using a continuous density evolution and differential evolution method. The method of repeating LDPC codes is proposed as well, which further improves the efficiency of multidimensional data coordination, effectively reduces the convergence signal-to-noise ratio threshold and extends the secure range of communication distance. The simulation results indicate that when the block length is 106, the convergence signal-to-noise ratio can be less than -6 dB, and the data reconciliation efficiency can achieve 90.27%. The amount of the extracted secure secret key is about 0.22 kb/s, and the secure range of communication distance exceeds to 80 km, which means the distance can be effectively extended in CVQKD system.

By the Runge-Kutta algorithm, the density matrix equation, which describes the interaction between laser and quantum wells, is numerically solved and the coherent control of population transfer in asymmetric semiconductor double quantum wells is studied. It is shown that the population transfer between two lower states can be controlled by adjusting the laser intensity and time delay among laser pulses. The relative phase among laser fields can be used to exactly control the population transfer probability. These results have potential application value in the fields of coherent superposition state preparation, quantum entanglement, and quantum information processing, and so on.

In order to address the critical compensation issue for attitude jitter in high-resolution (HR) images, a steady-state imaging model based on high-frequency angular displacement is proposed to detect attitude jitter and internal geometric accuracy compensation of HR satellite, and the accuracy of geometric calibration field is validated using digital orthogragh images/digital elevation model reference data. In the case of HR optical satellite launched in 2015, three experiments are carried out as follows: platform jitter detection, convergence analysis of attitude processing filter, and verification of jitter compensating effect. The compensation results show that the proposed platform jitter compensation model can effectively improve the imaging quality, and the internal geometric accuracy of jitter image reaches 1.5 pixel after compensation.

In order to improve the speed and accuracy of spectral calibration for echelle spectrometers, an automatic spectral calibration method by multiple characteristic wavelengths of mercury lamp is designed based on the spectrum retrieval algorithm. Spectral calibration experiments are conducted with a mercury lamp as the calibration light source. The results show that the spectral retrieval model is automatically corrected as long as the algorithm is within the range of the spectrum deviation. Choosing more calibration wavelengths and more uniformly distributed calibration wavelengths results in higher calibration accuracy. For the echelle spectrometer with wavelength range from 250 nm to 600 nm, choosing more than five calibration wavelengths will make the calibration accuracy be 0.01 nm, which is the theoretical resolution of the spectrometer. The automatic spectral calibration method makes echelle spectrometers more practical and valuable in engineering applications.

To solve the problem that common iterative pursuit algorithm has low success rate and precision of spectrum reconstruction in low sampling rate, an improve atom selection method of the spectrum reconstruction in low sampling rate is proposed. The spectrum dictionaries are constructed with AVIRIS and ROSIS hyperspectral data and the compressive sensing hyperspectral reconstruction experiment is conducted. The spectrum reconstruction precision, the sparse component extraction ability, and the success rate and accuracy rate in spectral reconstruction are analyzed in different views, respectively. Experimental results show that the proposed method is much better than conventional matching pursuit algorithms and also superior to the well know high precision method such as FOCUSS and MSBL algorithms.

Monolayer Mo films with the same thickness are deposited on the Si substrates by direct current pulse magnetron sputtering method. The influence of different deposition rates on the microstructure, morphology and optical properties of Mo films is investigated with stylus surface profiling system, X-ray diffractometer, field emission scanning electron microscopy, atomic force microscopy and ultraviolet-visible spectrophotometer, respectively. The Mo film thickness fitted by grazing incidence X-ray reflective spectra agrees well with the designed thickness, which indicates that the current process technology of Mo film is mature and stable. The morphology evolution of Mo films is affected by the deposition rate via altering the growth model and nucleation rate of Mo films. As the increase of deposition rate, the density, diffractive intensity of Mo (110) and average grain size of films increase, whereas the surface roughness decreases first and then increases.

Aim at the problem of rapid and high precise calibration in stereovision system, a calibration method for stereovision system based on solid circle target is proposed. The Canny-Zernike combination algorithm is used to locate the sub-pixel coordinates of circle center, and a sorting method for circle center based on a triangle target is given in detail. To reduce the computational complexity, the Levenberg-Marquardt (LM) algorithm is utilized to separately optimize the left and right camera for decreasing the number of optimal variables by half. The distance between two circle centers on diagonal direction of calibration board is regarded as the constraint, and the external parameters of cameras are optimized with LM algorithm. The calibration software for stereovision system based on solid circle target is complied, and the testing is performed in lab. Result shows that the proposed method has high precise, can achieve the automatic calibration and apply to engineering.

Waveform sampling lidar is the technical tendency in lidar technology in virtue of its high precision of three-dimensional geodesy and multi-layer target detection ability. Especially, array detection imaging lidar systems applying the streak tube principle have promising application prospect because of their full waveform sampling and high degree of integration. A new lidar system which is based on full waveform echo sampling and array detection of the streak tube is established to extract noise in echo signal of airborne flight experiments and summarize statistical properties of echo noise. A new denoising method for echo signal in waveform sampling lidar is proposed by comparing the different statistical properties between noise and echo signals.

An inversion algorithm based on the global search (GS) is proposed for the inversion of particle size distribution (PSD) from image dynamic light scattering (IDLS) data by image correlation. The inversion for unimodal particles and bimodal particles is simulated. The peak value and particle distribution index (PDI) of unimodal particles are 79 nm and 10%, respectively. The peak values of bimodal particles are 79 nm and 352 nm with PDI of 10%. It is shown that GS inverts PSD for IDLS data successfully. Based on the simulation, the experimental inversion of unimodal particles with peak value of 79 nm and bimodal particles with peak values of 79 nm and 352 nm, 79 nm and 482 nm is performed. Compared with the cumulants method for unimodal particles and the double exponential method for bimodal particles, GS algorithm can invert PSDs of unimodal and bimodal particles successfully, which indicates that GS is an effective inversion algorithm.

X-ray computed tomography (CT) has been widely applied as a powerful tool in both clinical medicine and industrial non-destructive test areas. With the development of flat panel detector in detection efficiency and anti-radiation property, the cone-beam CT based on the cone-beam source and the flat panel detector has been broadly used. However, the field of view of cone-beam CT is restricted by the hardware conditions such as detector size, and thereby the accessibility for larger size object applications is limited. In order to enlarge the field of view of cone-beam CT, a one-sided two-helical-scan single-slice-rebinning reconstruction algorithm is proposed. By performing one-sided two helical scans and normalizing the imaging geometries, two groups of helical cone-beam projections are recombined into one multi-slice parallel-beam projection, whose symmetrical feature is utilized to remove the transverse truncation in the data obtained, then the parallel-beam projection data covering the complete object cross-section is obtained, and thus the images are reconstructed by back projection. The experimental results show that the proposed algorithm can extend the field of view by 2.56 times with comparable imaging quality compared to the traditional algorithm.