By means of measuring the surface reflectance by an automated test-site radiometer (ATR), measuring the atmospheric parameters by CE318 sun-photometer, and measuring the diffuser-to-globe factor by the diffuser-to-globe irradiance meter, a hyperspectral surface reflectance at the time of overpass is determined. Based on the reflectance-based method, absolute radiometric calibration experiments with in-site automated observation at visible to near-infrared bands 1-7 of the moderate resolution imaging spectroradiometer (MODIS)/Aqua and MODIS/Terra are performed. The results show that the relative deviation in top-of-atmosphere (TOA) spectral radiances is less than 4% obtained from the in-site automated observation radiometric calibration (AORC) method and from MODIS/Aqua and MODIS/Terra, which indicates the in-site AORC method is at the same level of calibration accuracy with the manual calibration methods. The feasibility of the in-site AORC method is testified.

In the process of synchronous ground observation for quantitative remote sensing inversion, measurement uncertainty factors like human subjective factor, environmental change and condition restriction will induce data noise inevitably, which degrades the retrieval accuracy of the suspended matter concentration. Therefore, a measurement uncertainty-aware retrieval method named as the adaptive sample consensus extreme learning machine (ASAC-ELM) is proposed. ASAC-ELM integrates the merits of extreme learning machine (ELM), random sample consensus (RANSAC) and N adjacent points sample consensus (NAPSAC). The algorithm adaptively selects RANSAC or NAPSAC to estimate model parameters with the guidance of the parameter dimension, which avoids the problem that the ELM algorithm is sensitive to the non-zero normal distributed data noise. The ASAC-ELM algorithm selects inlying points (non-noise points) for model construction, thus can remove the interference from noise, and enhance the accuracy and flexibility of the model. In order to investigate the effectiveness of the proposed method under different noise conditions, a series of additive noise with non-zero mean normal distribution is introduced in the training data. The comparison among ASAC-ELM, ELM and traditional back propagation (BP) neural network algorithms is also conducted. The results show that for the retrieval of inland water suspended matter concentration under various noise conditions, the inversion accuracy and stability of ASAC-ELM is higher than those of ELM and the traditional BP neural network.

In order to improve observation precision of the meteorological visibility under non-ideal conditions, the equivalent visual threshold and general Koschmieder′s law are put forward and demonstrated based on the atmospheric optical transmission theory. Under the non-ideal condition, the observation visibility formulas of relative error and the correction coefficient are established. The weather visibility relative error and the correction coefficient of the non-black targets are analyzed under the conditions of non-parallel level of the sky background. Results show that the correction coefficient of visibility observation is influenced by target reflectivity, background reflectivity, the distance between target and background and atmospheric extinction coefficient, and the distribution of correction coefficient has a certain regularity. In order to meet the needs of visibility observation business, the results of visibility observation under non-ideal conditions need to be modified, especially when the correction coefficient is less than 0.8.

Time domain crosstalk in fiber Bragg grating Fabry-Pérot (F-P) interferometric hydrophone array system limits its practical application seriously. So it is necessary to analyze the generating mechanism of time domain crosstalk. With the phase generator carrier (PGC) demodulation, the effect of amplitude on phase is discussed for the interference signal. Then two comparative hydrophone systems with different reflectivities of fiber Bragg grating are fabricated and tested. Experimental results show that amplitude of the crosstalk is high for the demodulation signal with high reflectivity of grating (R=0.6), because the phase of the crosstalk signal hardly reduces on the basis of the PGC demodulation. On the other hand, the crosstalk has about 34 dB suppression with the low reflectivity of grating (R=0.02). Furthermore, the crosstalk in low reflectivity grating system is verified to be composite of the Doppler noise of precedent hydrophones and time domain crosstalk.

tIn the atmospheric channel, the spatial correlation among the beams of wireless optical multiple input multiple output (MIMO) system caused by the antenna spacing and the channel fading can seriously affect the wireless optical communication system performance. We proposed a channel model for the correlative wireless optical MIMO channel based on the pulse position modulation (PPM), and the average channel capacity in lognormal fading channel is derived by using the Wilkinson method. The simulation results indicate that the average channel capacity is decreased due to the spatial correlation, and the effect is enhanced with the growth of the total energy. In addition, in a completely correlative channel, the impact of transmitter correlation on the channel capacity is more obvious than that of the receiver correlation.

A high sensitivity magnetic field and temperature sensing structure based on photonic crystal fiber(PCF) is designed. Magnetic liquid is filled in an air hole of PCF cladding to form the directional coupling structure. The magnetic field and temperature variations of the structure are measured. The characteristics of the sensor are studied simulatively by using a full vector finite element method (FEM). The results show that the sensing structure can realize measurement within the magnetic field range of 100~250 Oe（1 Oe=79.58 A/m） and the temperature range of 10~60 ℃. The magnetic field and temperature sensitivities can reach 1.10 nm/Oe and -3.86 nm/℃ within the range.

In the multiple-input multiple-output (MIMO) visible light communication (VLC) system, the effect of multipath can not be ignored specially at high bit rates. A modeling method for indoor multipath channel in VLC-MIMO system is proposed when there is the time dispersion at transmitter. A modeling method for indoor multipath channel in VLC-MIMO is proposed. Because the orthogonal frequency division multiplex (OFDM) can effectively resist inter-symbol-interference (ISI), the MIMO-asymmetrically clipped optical (ACO)-OFDM system is proposed by combing ACO-OFDM and MIMO techniques. The impulse response of each pair of LED and photoelectric detector（PD） is calculated by using recursive method. Then considering that the link between the centroid of LED and PD array is the equivalent line of sight (LOS) link, the multipath channel is established. The analysis results indicate that the gain of multipath channel is large and the multipath interference is weak when the PD arrays are at the center of the room, while the signal attenuation is large and the multipath interference is strong when the PD arrays are in a corner. The Monte Carlo bit-error-ratio (BER) simulation of MIMO-ACO-OFDM system are derived and modeled during both zero-forcing (ZF) and minimum mean square error (MMSE) signal detection. The BER performance degrades as the modulation order increases. When the modulation order is less than 64, the BER performance of ZF is worse than that of MMSE and the performance difference between ZF and MMSE decreases as the SNR increases. When the semi-angle at half power of LED becomes small, it has a better performance at the center of the room due to the strong beam convergence. Moreover, when the field of view (FOV) of LED becomes small, it has a better performance at the center of the room due to the decreased multipath interference, although the received power remains the same.

A high order coupled nonlinear Schrdinger equation suitable for ultra-short Airy pulses is solved by means of the split-step Fourier transform method, and the evolution of the interaction of ultra-short Airy pulses propagating in fibers is numerically simulated by Matlab software. The results show that the negative third-order dispersion effect accelerates the penetration of wave packets and makes ultra-short pulses propagate over a long distance. In contrast, the positive third-order dispersion effect slows down the propagation of ultra-short pulses, and if the third-order dispersion coefficient is large enough, the pulse oscillation is transferred from the leading edge to the trailing edge. The self-steepening effect makes ultra-short pulses generate time-domain shift in the form of soliton splitting. The intra-pulse Raman scattering effect results in a Raman self-frequency shift at the long wavelength side of the pulse, so that it can transfer the pulse energy from the leading edge to the trailing edge. Under the combined impact of the self-steepening effect and the intra-pulse Raman scattering effect, an ultra-short pulse shows a time-domain displacement and the energy at its leading edge is transferred to its trailing edge. The simultaneous existence of three effects of third-order dispersion, self-steepening and intra-pulse Raman scattering strongly influences the self-bending and self-acceleration characteristics possessed by the interaction between ultra-short Airy pulses.

Fresnel lens with large aperture is stitched by multiple segmented mirrors, and miss-adjustment error of segmented mirror will affect imaging quality of stitched Fresnel lens. Using the Rayleigh criterion as evaluation criteria, according to Fresnel lens wave-front aberration theory, each degree of freedom allowed limit for Fresnel lens miss-adjustment error tolerance is analyzed, and the theoretical formula is given. Analysis result shows the bigger the stitched Fresnel lens F-number is, the looser each freedom degree of the miss-adjustment error tolerance limit becomes. Using Zemax software to simulate, the stitched Fresnel lens wave-front is tested by interferometer. The result shows that the maximum error between the stitched wave-front and the theoretical calculation is 0.006λ, which verifies the correctness of the theory. The results provide a theoretical basis and guidance for design, test and alignment of the stitched Fresnel lens.

The spectral information of interferogram is obtained by Fourier transforms with spatial heterodyne spectrometer (SHS). So the spectrum signal to noise ratio and spectrum quality are directly influenced by interferogram modulation efficiency. In view of the influence factors of interference fringe modulation is carried on the theory of quantitative analysis, including splitting characteristics of beam splitter, aberrations of optical system (surface wavefront error of interferometer elements, output wavefront error of collimating lens and modulation transfer function curve of imaging lens), the extended source and pixel size of detector. Experimental result show that the measured result of recovering spectra by the SHS instrument is in agreement with the theoretical calculation. Quantitative analysis of the interference fringe modulation for interferometric spectrometer provides theory basis for the design of each function module.

The vertical optical measurement technique with coaxial projection and imaging optical axis based on the modulation analysis provides a means to solve the problem of shadow and occlusion for measuring the complex surface or step-like surface. Without the phase truncation calculation and phase unwrapping steps, only the modulation information is needed to reconstruct the surface of the tested object in the method, which is widely used as one of the optical three-dimensional measuring techniques. Aiming at the problem of the vertical optical measurement based on the Fourier transform analysis, this paper introduces wavelet transform in modulation measurement profilometry by using partial analysis and multi-resolution characteristics of the wavelet transform for improving the measurement precision based on single fringe analysis at every scanning position. The relationship between the wavelet coefficient and the modulation is deduced, and the computer simulation and the experiment show the validity of the method.

Based on the vector aberration theory, a method to analyze aberration field distribution induced by the Fringe Zernike polynomial off-axis freeform surface is proposed. The analytical expressions of third-order astigmatism and coma of off-axis freeform surface system are deduced. The aberration nodal distribution of off-axis system caused by the freeform surface is analyzed. According to the aberration distribution properties, with proper Zernike terms selected and optimized, an off-axis two-mirror freeform surface long-wave infrared optical system is designed, and the system has an effective focal length of 500 mm and a pupil diameter of 300 mm. The system is compatible with the uncooled infrared detector that has a format of 384 pixel×288 pixel, and the pixel size is 25 μm. The third-order astigmatism and coma nodes of the optimized system are both moved into the field of view, the asymmetric aberration induced by off-axis is balanced, and the imaging quality of the system is close to the diffraction limit that meets application requirements.

Digital micro-mirror device (DMD) has the advantages of high refresh rate and high diffraction efficiency, and it is an ideal dynamic computer-generated hologram loader. Light-emitting diode (LED) is chosen as light source to replace the laser, since it can effectively reduce the speckle noise caused by both temporal and spatial coherence of laser. A holographic display system using LED as light source and DMD as computer-generated hologram loader is proposed in this paper. The modulation characteristics of DMD as a hologram display device are studied and the reason why worse image quality caused by LED chosen as represention light source is analyzed. The impact of coherence and collimation of LED on reconstruction image quality is analyzed. Numerical simulation results verify the above analysis. According to the results, a color filter and a spatial filter are used to improve the coherence of LED and an aspheric collimation lens is used to improve the optical efficiency. The experimental results show that the system has good display quality.

Based on the free-space method, a set of D-band transmittance measurement device is designed and developed, and typical clothes, shoes and wrappers are measured by the device. The device consists of a D-band vector network analyzer, two Gauss beam antennas, two hyperbolic lenses, an optical guide and a clamp. The through-reflect-line (TRL) calibration method and the time domain gate technique are adopted to improve the measurement accuracy and guarantee the validity of the data. Ten kinds of clothes with different thickness are measured by the device. The trend of data agrees with the published data well. Furthermore, three kinds of shoes and six kinds of wrappers are measured. The measured results are similar to those obtained by the security system, and are helpful to the design of security systems.

The digital phase conjugation technology is used to investigate the optical field recovery characteristics in scattering media. The scattering medium is treated as a plane of phase and amplitude random modulation, the optical field recovery effect under three different modulation conditions is deduced theoretically, and a numerical simulation is performed. The results show that optical fields can be ideally recovered for the pure phase modulation or the Gaussian amplitude random phase modulation with a Gaussian curve wide enough. In pure phase modulation, higher wavefront recording precision and larger aperture size will result in better recovered optical fields. The optical field recovering can accommodate some noise, and the intensity recovering is more robust to noise than the phase recovering.

High dynamic range imaging technique can avoid the brightness loss and aberration of the images because of shooting direction such as backlight and different exposures, which will affect the information collection of real scene. The technique is in favor of higher quality imaging in the complex environment and widely used in pattern recognition, intelligent transportation system, remote sensing, military reconnaissance and other research areas. Camera response function calibration is the key to the high dynamic range imaging technique, which can establish the mapping relation between real scene irradiance and acquired image brightness values. Thus the high dynamic range images in actual scene are acquired. The proposed algorithm can acquire the camera response function of the imaging system by single frame input image, and the calibration efficiency is enhanced greatly. Which is suitable for the under-exposure and over-exposure images, and expands the application range of camera response function calibration.

On-orbit modulation transfer function (MTF) of the space camera usually contains the MTF with atmospheric effect, and therefore it does not objectively reflect the imaging performance of the space camera. Based on the relationship among the modulation depth of the object target, the object target before the entrance pupil of space camera and the image, a novel technique is proposed to exactly test the atmospheric MTF and the on-orbit MTF without atmospheric effect. The test results of a space camera with the above measuring technique indicate that the real time atmospheric MTF is 0.762, the on-orbit MTF with atmosphere effect is 0.177, and the on-orbit MTF without atmospheric effect is 0.232. The relative error between the MTF value measured in the laboratory and the on-orbit MTF without atmospheric effect is 4.7%.

The basic principles of wavefront sensing technology based on phase diversity (PD) are discussed. The noise adaptability of the PD technique and the influence of the detector defocus distances error on the accuracy are analyzed. PD technology is applied to assisted align off-axis and coaxial three-mirror reflecting optical system, and the wavefront maps of multiple fields of view are compared with interferometric test results. Experiment results show that wavefront sensing and assisted alignment of three-mirror reflecting optical system can be realized by using PD technology. Compared with the results of interferometric test, results of wavefront sensing are all better than 0.02λ RMS. The accuracy can satisfy the requirement of engineering.

In order to improve the speed and accuracy of generating inverse fringes, a method of generating inverse fringes based on Delaunay triangulation is proposed. With this method, the forward mapping relation between projector coordinate and camera coordinate is used. The phase points on camera coordinate compose a set of scattered phase points, with which the Delaunay triangulation mesh is generated. Meanwhile, the phase points on projector coordinate are treated as interpolation points, which are used to find the corresponding Delaunay triangulation. After the interpolation operation is finished, the expected phase values of inverse fringes along vertical and horizontal directions are obtained. With these phase values, the inverse fringes can be further generated. The computer simulation and real object simulation experiments both demonstrate that the above method possesses advantages to improve the accuracy and speed of generating inverse fringes, which makes it has a high application value.

It is a long-term goal and of great significance to seek a new method towards direct realization of the SI unit of sound pressure and no longer rely on the sensitivity of standard laboratory microphone in the field of airborne acoustic metrology. Laser Doppler velocimetry technique can be used to realize the goal. In this paper, the optical interferometer without Doppler frequency bias is designed to investigate the realization of sound pressure in traveling wave tube with the photon correlation spectroscopy (PCS). According to the linear relationship, the sound pressure of plane acoustic wave can be acquired by the demodulated particle vibration velocity from the Doppler signal. The experiments are also implemented to demonstrate the feasibility and accuracy of this scheme with the working standard microphone as reference. For the acoustic frequency of 315 Hz, the deviation between the two methods is less than 0.5 dB with the measured sound pressure level varied every 1 dB in the range from 100 dB to 110 dB. For the measured sound pressure level of 105 dB, the deviation is less than 0.3 dB with acoustic frequency of 315, 400, 500, and 800 Hz.

The scheme of Nd∶YAG and Nd∶YVO4 crystal composite application is proposed. The Nd∶YAG crystal with good thermal conductivity and optical properties is used as the front end of pump light absorption crystal, and placing a wider absorption spectrum Nd∶YVO4 crystal at its rear-end to absorb the pumping light energy composition, which is not absorbed by Nd∶YAG crystal due to the spectral width does not match. Two kinds of crystals in the 1064 nm wavelength emission spectrum overlap each other. The absorption of pump light energy can be converted to common oscillation laser wavelength, the efficiency of pumping light is improved. The proposed composite applications can effectively reduce the oscillating light power changing with pumping source working temperature fluctuations. Experimental results show that the proposed composite application scheme improves the optical-to-optical efficiency of pump sources by 22.9%, and the sensitivity of output power affected by temperature is reduced from 7% to 1%, comparing with the single Nd∶YAG crystal scheme.

Femtosecond lasers focused on metal surfaces can induce subwavelength periodic structures, which can make metal surfaces colorize. The hue and brightness of the structural colors vary with the incident angle of illumination and sample rotation angle, which indicates the structural color has directional property. In order to investigate the uniformity of femtosecond laser colorization on metal surfaces, based on the color system of International Commission on Illumination the color deviation is analyzed. The results show that the larger the processing area is, the worse the color uniformity is. With the specified experimental parameters, when the processing area of sample reaches to 64 mm2, the difference between edge color and center color is obvious to human eyes. Two colorized icons are made to demonstrate the effects of the directionality and uniformity of structural colorization on the practical application of femtosecond laser color marking.

For the vision measurement of aero large parts and components, traditional camera calibration methods relies on the high precision of standard targets which are difficult to be produced. The precision of these methods can not meet the need of large field of view for traditional camera calibration. Aiming at these problems, a site-calibration method based on collinear constraints of mark points located at the four corners is put forward. Some calibration control points are arranged close to the principal point within space measurement field of view, and the calibration initial results can be solved using the linear transformation. The collinear constraint rulers are placed in the field of four corners, the distortion coefficients can be calibrated using cross-ratio invariability and linear fitting. The number of control points and the whole parameters of calibration can be optimized, and the optimal site-calibration result can be obtained. The relative calibration experimental results show that in the field of view of 2.5 m×1.8 m, the reconstruction error of site is less than 0.07%, which meets the demand that the high precision of site-calibration, and has good robustness and adaptability for the vision measurement of aero large parts.

In order to recover depth information from monocular infrared image, a depth estimation algorithm based on novel deep convolutional neural networks (DCNN) is proposed. The texture energy and texture gradient of infrared images are extracted by using Laws′ masks and the gradient detector at different scales. These two types of texture information are considered as the first kind of features. The selected gray values and their statistical histogram in specific areas are considered as another two kinds of features. The DCNN are trained on these three kinds of features with the corresponding depth labels respectively. The trained DCNN are then utilized to estimate the depths of testing monocular infrared images respectively. Experimental results show that compared with other methods, the DCNN trained by texture information can estimate the depth much better than those of the existing methods, especially in the depth changes of local scenes.

Boron-doped diamond thin films are homoepitaxially grown on 3 mm×3 mm×1 mm commercial Ib(100) diamond substrate with high temperature and high pressure(HPHT) by microwave plasma chemical vapor deposition(MPCVD). The magnetron sputtering and electron beam evaporation techniques are used to prepare the diamond Schottky diodes with different structural parameters. The test results indicate that the surface of the growed diamond thin films are very smooth and obvious atomic steps can be observed. The device has obvious rectifier features. The forward conduction resistance of the device is 20 Ω at 300 K when the diameter of Schottky electrode is 100 μm, the distance between the Schottky electrode and ohm electrode is 10 μm and the applied voltage is -15 V. The reverse saturation current is about 10-6 A and the reverse breakdown voltage is about 103.5 V. The greater the distance between the Schottky electrodes and ohm electrode is, the higher the reverse breakdown voltage is and the smaller the forward current is.

The electronic structure and optical properties of Sn-doped ZnO for different doping concentrations are calculated by using the first-principles under the framework of density functional theory with the generalized gradient approximation and the Perdew-Burke-Ernzerdorf functions. The effect of doping concentration on the crystal structure, band structure, density of state, and optical properties is studied. Meanwhile, according to the calculated band structure and charge density of difference, the effect of doping site on the calculated results is investigated. The results show that with the increasing Sn doping concentration, the ratio of lattice constants c to a is stable, and the doped structure does not distort. The total energy of the doped system increases gradually, thus the stability weakens, and the band gap decreases first and then increases. The doped SZO (Sn∶ZnO) system becomes an indirect band gap semiconductor, and a large number of conductive carriers, which are contributed by the doped Sn atoms, are introduced to the bottom of the conduction band. As a result, the conductivity is significantly improved. Moreover, a V-shaped curve occurs between the Fermi level and the top of the valence band, which shows the characters of half-fill state. After doping, density of state peak of the valence band moves to the lower energy by about 1.5 eV. The donated and received electrons in the same-layer doping are more than those in the neighbor-layer doping and the alternate-layer doping whereas the former band gap is smaller than the latter. The absorption edge has a red shift, and the ultraviolet absorption capacity is enhanced significantly. The imaginary part of the dielectric function increases, and the primary transition peaks shift to higher energy. The calculated results show that SZO (Sn∶ZnO) is a good transparent conductive film.

Image reconstruction with large scale of fluorescence molecular tomography (FMT) data needs a large amount of computational memory and time. In order to reduce the ill-posedness of FMT and speed up the reconstruction，an accelerated reconstruction method for FMT based on locality preserving projections (LPP) and sparse regularization algorithm is presented, combining with the manifold learning and compressive sensing theory. The original fluorescent multi-projection data is reconstructed. Simulation experiment of non-homogeneous cylinder single and double goal and real experiment are performed to evaluate the reconstruction effect and time of the proposed method respectively. Experimental results demonstrate that the proposed method can ensure the accuracy and resolution of the FMT reconstruction image and reduce the reconstruction time greatly.

Photonic lattices are optically induced in photorefractive crystals and the transmission of dipole solitons in photonic lattices is studied. By means of numerical simulation, when the dipole beams are launched at the diagonal lattice sites, different transmission results of dipole beams are obtained when the intensity of bias field voltage, the intensity of beams and the depth of lattices are changed. These results show that the input dipole beams can overcome the interaction between the two humps as a result of the existence of the lattices. Under suitable conditions, out-of-phase dipole solitons exist and are stable. In contrast, in-phase dipole solitons are always instable, and the instability is weaker if the beam intensity is lower.

Space-based multi-model differential optical absorption hyperspectral imager is a new type space atmospheric sounder. It requires the multi-model sounding functions including nadir, limb and sun occultation. It chiefly sounds the trace gases such as SO2, NO2 and so on. A new method of multi-model sounding with high spectral resolution is proposed. The method uses two-scanning mirrors to switch differential sounding model, double-spectrometers to reduce the spectral stray light, and dichroic filter to divide working waveband into three channels. An optical system of multi-model differential optical absorption hyperspectral imager is designed. The instantaneous field of view is 1.8°×0.04°. The F number of the system is 2. The working waveband is 250~500 nm, which is divided into three channels, the first one is 250~310 nm, the second one is 300~410 nm, and the last one is 400~500 nm. Optimization design and analysis are performed by ZEMAX-EE software. The spectral resolution is 0.12 nm in the waveband of 250~310 nm, which satisfies the requirement specification of no more than 0.4 nm. The spectral resolutions are 0.25 nm and 0.23 nm in the waveband of 300~410 nm and 400~500 nm, respectively, which satisfy the specification requirement of no more than 0.6 nm. The modulation transfer function (MTF) of multi-model differential optical absorption hyperspectral imager is more than 0.98 at characteristic frequency of 0.25 lp/mm in the spatial dimension. The design results satisfy the requirements of multi-model differential optical absorption hyperspectral atmospheric sounding.

An integrated simulation method is established, and the performance assessment of optical systems is obtained, accordingly. New assessment functions are used for rigid-body displacement separation, which is conducted by deformation data from the Ansys thermal simulation. The computational accuracy of 0.3% is tested through setting displacement parameter and random value. Householder algorithm is adopted for Zernike fitting, and the Zernike coefficients are used as the interface for data exchange with Zemax, which is achieved by dynamic data exchange technology. The simulation of an objective lens is taken as an example, and a distortion-thermal curve is presented. The result indicates that in order to guarantee measuring accuracy of 1″, the working temperature should be in the range from 14 ℃ to 26 ℃.

A new method of laser beam orientation calibration based on deflection and translation of optical path is proposed. In order to realize automatic calibration of optical path and stable control of laser beam orientation, a control mechanism with a combination of rotation and translation is adopted to carry out feedback deviation adjustment of laser beams. Two detectors are used to measure the direction and position of laser beams. Via a simple design of optical path calibration, the independent feedback adjustment of angular deviation and parallel deviation along the horizontal and vertical directions can be realized, and the coupling among different deviation in the adjustment process can be avoided. The introduced parallel deviation control is not sensitive to propagation distance, and therefore it is beneficial to enhancing the adjustment stability and accuracy. Moreover, the laser beam orientation calibration system based on this method is set up. The experimental results show that with this system the expected calibration effect can be achieved.

For hard and brittle aspheric optical mirror with characteristics of high hardness, special profile, large processing area and high processing quality requirement, a generating grinding method for aspheric surface based on the normal tracking and grinding with cylindrical grinding wheel end is developed. Based on this, intelligent real-time detection and adaptive control method are designed. Applying these methods, the precise control of the grinding motion track, automatic compensation of grinding wheel wear,automatic control of grinding pressure and automatic adjustment of process parameters are realized. Experimental results show that the method for generating precision grinding of aspheric surfaces which is grinded with cylindrical grinding wheel end and based on normal tracking can ensure the accuracy of the aspheric surfaces. The real time detection of grinding wheel wearing and adaptive control method can improve the machining quality and efficiency.

For a Cook optical structure with small F number and short back working distance, a channel splitter based on the optical free-form surface and the corresponding test solution are proposed. The solution uses multi-degree of freedom of optical free-form surface to correct the astigmatism caused by tilt of the splitter, which makes transflective two-channel optical image quality approach to the diffraction limit. Computer-generated hologram testing technology is proposed to test the free-form surface of the splitter. The peak valley of the wavefront is 0.0096λ，the root mean square of the wavefront is 0.0017λ in the testing system. The splitter and its testing solution meet the design requirements, and provide practical reference for the optical system with small F number and short back working distance.

There is a race to develop spaceborne high-resolution video cameras since Skysat-1 is launched successfully. To reduce manufacture cost and adapt to small satellite platform, it is urgent to design and develop a long focal length optical system with small volume, light weight, easy implementation and imageable for two-dimensional field of view. The full field of view coaxial three-mirror anastigmat (CTMA) appropriate for two-dimensional field of view imaging is studied. The problems of its inherent secondary obscuration and imaging plane extraction difficulty are solved. The conditions of eliminating the secondary obscuration and the relationship between the total length of the system and the tertiary mirror relay imaging magnification are presented based on paraxial geometry optics. Then the reasonable tertiary mirror relay imaging magnification and other parameters are discussed and obtained. A flat mirror near its exit pupil can be used to deflect the image plane without obscuration. The flat mirror group is used to fold optical path and the total length of the system is decreased. The method and the result for determining the initial structural parameters are presented. The optimized optical system without secondary obscuration is designed, whose effective focal length (EFL), field of view and F number are 10 m, 1.1°×1°, 14.3, respectively. The total length of the system is 1/8 of EFL. Its imaging quality is near diffraction limit.

As the cross dispersion, the two-dimensional spectrum of echelle spectrometer cannot calibrate the wavelength spectrum directly, therefore the spectrum reduction model for the transmission prism echelle spectrometer with the C-T structure is built. The dispersion rule and relationship between prism and echelle is analyzed, and the relational expression of wavelength and image coordinates is also established. According to the optical structure and optical transmission characteristics, the deviation of coordinates caused by the element can be regulated, and the accurate image coordinates can be calculated. Finally the spectrum reduction model is accomplished. This model can recover the two-dimensional spectrum of echelle spectrometer quickly and accurately, and calibrate the wavelength. The error of image coordinates calculated by the model is less than one pixel.Key words

Based on the dielectric loaded graphene plasmon waveguide the double dielectric loaded graphene surface plasmon polariton waveguides (DDLGSPPW) is proposed using two refractive index dielectrics. The change rule of effective refractive indices of graphene surface plasmon polariton (GSPP) modes versus incident wavelength, width, height and dielectric constant of dielectric 2 in DDLGSPPW are studied, using the effective index method and the finite element method. The results show that the characters of the obtained and the four layer dielectric slab waveguide are similar, through changing the effective refractive indices and the modulus with the parameter of dielectric 2, and adjusting the distribution of the two dielectrics mode. The characters of DDLGSPPW may have new application value in the integrated optical devices.

A quantum ranging scheme based on two-mode squeezing and homodyne detector is proposed to avoid the difficulties of generating and maintaining of quantum states in multi-photon entangled ranging schemes and optical-delay-caused measuring range blocking in quantum correlation distance ranging schemes. Using the character that the two-mode squeezing beams has the largest cross-correlation of the corresponding quadrature of the two modes when the synchronization, the reference photocurrent is delayed to match the signal photocurrent in order to acquire the distance ranging. On the basis of double balanced homodyne detector to detect the quadrature of entangled light beam, the methods of correlation function estimation, correlation noise estimation and correlation matrix analysis are proposed to estimate the delay time, respectively. The theoretical demonstration and simulation verifications show that the proposed quantum ranging scheme has certain feasibility and practical significance.

Determination and calibration of high precision sensor line of sight (LOS) attitude are the key prerequisite of high precision tracking and location of targets in space based optical surveillance systems. LOS determination and calibration of staring sensor with high frame frequency and narrow field of view are difficult points of the problem. Based on the research of the imaging model and observation characteristics of staring sensor, the real time LOS attitude determination and calibration algorithm using landmark control point is proposed. The influential factors (including thermal distortions error, assemble error, and so on) of staring sensor LOS attitude error are equivalent to bias angle of LOS attitude. By establishing the observation equation of landmark control point and the state transition mode of bias angle, and using an extend Kalman filter (EKF), the real time estimation of bias angle and the high precision LOS attitude determination and calibration are achieved. The simulation results show that the precision and timeliness of the proposed algorithm meet the request of target tracking and location process in space based infrared surveillance system.

To investigate the influence of smoke concentration on the transmission characteristics of polarized light in the smoke environment, carbonaceous particles are used as the research object to study the variation in the degree of polarization (DOP) of linear polarized light in the horizontal, vertical, +45° directions, and right- and left-hand circular polarized light after transmission in different medium concentrations. The Monte Carlo method is used to build a simulation model and analyze the polarization transmission theory. According to the uncontrollable factors for the concentration in actual measurement, the optical depth is applied as the description of the concentration, and the influence of concentration on the transmission characteristics of polarized light is verified by the simulation and the experiment. The results show that DOP of the incident polarized light decreases gradually with the increasing medium concentration. When the linear polarized light in horizontal, vertical and +45° directions is incident, the DOP variation tendency with the concentration is nearly the same as each other. DOP variation trend for the right- and left-hand circular polarized light basically coincides. DOP of the circular polarized light is close to that of the linear polarized light at low medium concentration, and is always higher than the latter at high medium concentration, which demonstrates better polarization-maintaining ability of circular polarized light under low visibility and high medium concentration.

Under a Gaussian beam incidence, the fundamental statistical moments of Gaussian beam scattering fields from targets with arbitrary shapes are derived. Taking an ellipsoid-shaped target as an example, the correlation function, covariance, and incoherent component ratio of the scattering fields in three different materials are studied by the numerical method. The variation of the incoherent component ratio of the ellipsoidal target with the scattering angle is numerically obtained. The results show that the incoherent component takes only a small proportion of the total scattering components. The target pose, surface material and its roughness have influence on the incoherent component ratio. The smoother the rough surface is, the smaller the incoherent component ratio is. The incoherent component ratio of metal material is smaller than that of non-metallic painting material. The incoherent component ratio of the scattered field from the scaled target has a similar distribution with the above, except for some difference in value.

In the waveband of 0.55～0.85 μm, how to modulate the phases while ensuring high transmittance of thin films is the key technology of multiple polarization imaging module project. Phase modulated antireflection coatings has been designed and fabricated on the substrate H-LaK4L(optical glass) by utilizing asymmetry equivalent layers theory and using three types of oxide coating materials: Ta2O5、Al2O3、SiO2. After process optimization, in the waveband of 0.55～0.85 μm and with the incidence angle of 28°, the results indicate that the average transmittance is greater than 97% while phase difference is less than 1°. The coatings can pass the environmental test required for space optical thin film products, and meet the reliability requirement of the project.