As channel selection acts an important influence on calibration accuracies in the calibration of imaging spectrometers by transfer radiometer, herein MATLAB language was adopted to compile a merit function, which was used as the evaluation standard of inversion accuracy of the spectral curve of a blackbody at a temperature of 3000k. The MATLAB genetic algorithm and the pattern search algorithm toolbox were employed to solute the minimal value of function that was the optimal spectral channel distribution of the transfer radiometer. Finally, influence factors of the number of spectral channels, including the shifting of center wavelength, the broadening of bandwidth and measurement system error, were analyzed. The results show that the spectral inversion accuracy of is 0.12%, when the shifting of center wavelength does not exceed 0.2 nm, the broadening of bandwidth is not more than 0.02 nm, and the system error is less than 0.1%.This study can guide the design of transfer radiometers and has a great significance to improve the calibration accuracy of earth imaging spectrometers.

In order to improve the measurement accuracy of dividing error of encoders, the principle and method to measure dividing error by using angular polygon and autocollimator were introduced, and the error sources were analyzed. According to the measurement principle, coordinate systems of angular polygon and autocollimator were established. Utilizing the method of coordinate transformation, precise mathematical models were deduced for indicating the relationships between dividing error and misalignment errors, such as parallelism error between the angular polygon working surface and the axis of the tested unit, perpendicularity error between autocollimator optical axis and the working surface of angular polygon, parallelism error between vertical wire of the autocollimator and the axis of the tested unit. In order to verify the error models of misadjustment, three experiments were performed in the laboratory, taking the positioning error of a single-axis position turntable as test object. The experimental and theoretical results have good consistency and the maximum deviation was less than 0.9″, which indicate that the error models of misadjustment are applicable to guide dividing error measurement.

In order to achieve rapid fabrication of hydrogel cell scaffolds, the computer-aided holographic method was employed into the traditional femtosecond laser two-photon processing. The generation of hologram as well as the influence of its parameters on the processed cylinder microstructure was studied. Firstly, Ring-shaped Bessel-beam holograms were generated according to the Bessel light wave equation and its transmission function, of which the parameters effect on the size and quality of ring structure was analyzed. Then the generated holograms were used to process tubes with a thickness of 0.8 μm and diameters from 8 μm to 15 μm in the polyethylene glycol double acrylate (PEGDA) hydrogel. Finally, an efficient and rapid processing of hydrogel cell scaffolds was realized experimentally and the diameter of hydrogel tubes was 8 μm. This technique can remarkably improve the efficiency of femtosecond laser processing, thus has great potential in many applications such as biological culture and chemical analyses.

An optical fiber Sagnac interference loop with two angle shift spliced polarization maintaining fibers was designed and fabricated to realize high-sensitivity temperature sensing. In the optical fiber Sagnac loop, two sections of polarization maintaining fibers were spliced with an intersection angle of 45° between their fast axes. The interference spectrum formula of the Sagnac interference loop by the Jones matrix was deduced theoretically. The influence of main parameters of the formula on the output characteristics of the Sagnac interference loop was studied in simulation. The simulation results show that vernier effect can be achieved in the optical fiber Sagnac interference loop with two angle shift spliced polarization maintaining fibers, of which the average length and the length difference affect wavelength intervals and envelope periods of output interference spectra of the Sagnac interference loop, respectively. In addition, the optical fiber Sagnac interference loop with two angle shift spliced polarization maintaining fibers was applied as a fiber temperature sensor experimentally. The results show that the sensitivity of the proposed temperature sensor based on Sagnac interference loop is -2.44 nm/℃ within 2 cm temperature zone, which is 14.97 times as compared with that (-0.163 nm/℃) of ordinary Sagnac loop sensors without vernier effect.

Aiming at the protection of reflective main optical system for shipboard photoelectric tracking systems, an 1 m size optical window and its frame were designed, manufactured and tested. Firstly, the performance and strength of the window material were analyzed. According to the influence of the difference between internal and external pressures, the weight and the rotational inertia force on the window, the minimum thickness was determined. Then, the effects of deformation of optical window and frame on optical path difference and the environmental adaptability were studied base on the index requirements. Moreover, the aplanatic non-planar grinding method was used for machining and inspecting the optical window. Finally, an fused silica optical window with 1 032 mm in diameter, 80 mm in thickness and 1 010 mm in aperture diameter were successfully developed. The optical path difference is RMS=0.062 8λ@632.8 nm. The results proves that the optical window can provide an effective protection for main optical system of photoelectric tracking systems.

In order to measure the thickness of large-sized hollow rotating parts, a non-contact measuring method based on double laser displacement sensors was proposed. With unknown relative position between rotating parts and the double laser displacement sensor, the measuring errors induced by the installed eccentricity of rotating pieces as well as the non-collinear relationship between dual-sensor optic axis and the origin were analyzed. Meanwhile, mathematical model between dual-sensor signals and the thickness of the measured surface position was deduced. Thereby, the actual thickness value can be extracted from the sensor signals with the aids of the phase difference measurement based on correlation theory, Newton iterative method and the circulation translation method. Through the simulating experiment for the thickness detection algorithm of the rotary body, it is indicated that the relative detection error remains within 0.5% when the amplitude of the interference component in the detected signal is no greater than 0.3mm, which can be used as a reference for rotating speed adjustment of the turntable. The analysis on the experimental measurement data shows that the maximum relative translation of two detection signals is 4; with eccentricity compensation and cycle-spinning of the data, the measurement repeatability error of the thickness detection algorithm is less than 0.05 mm, which meets the detection precision requirement large-sized hollow rotating parts in any position.

In order to improve the reliability and repeatability of phase counting and subdivision of quadrature signals for homodyne interferometers, a high accuracy error compensation algorithm for quadrature signals combined error correction with phase-based counting and subdivision was proposed based on the Heydemann correction model of quadrature signals and digital signal processing technology. Matrix manipulation based on the least square method was adopted to calculate compensation parameter initials of quadrature signals, which were corrected by iterative operation to further improve the compensation accuracy. Then phase-based counting and subdivision algorithm was established base on the revised signals. Finally, the algorithm was verified with Matlab software. Experimental results show that the proposed algorithm can accurately compensate the error of the quadrature signals, and the measurement accuracy achieves sub-nanometer or even picometer scales, thus the demodulation accuracy of the measured signals can be improved effectively.

To detect weak underwater acoustic signal over large areas, an optical Distributed Acoustic Sensing(DAS) scheme based on space difference of Rayleigh backscattering was presented. In this scheme, Rayleigh backscattered light with phase changes induced by the acoustic signal along a single-mode sensing fiber was split and fed into an imbalanced Michelson interferometer. Adjusting the path difference of the imbalanced Michelson interferometer, the Rayleigh backscattered light interference of different lengths of adjacent space segments along the sensing fiber was realized. Subsequently, the phase information including the acoustic signal was demodulated by the 3×3 coupler demodulation technology. An underwater acoustic wave measuring system based on DAS was implemented, which can not only locate the two acoustic positions accurately in real time, but also restore the amplitude, frequency and phase of sound waves. In addition, the acoustic phase sensitivity is -148.8 dB(re rad/μPa) at 1 kHz, and the frequency response flatness at frequencies ranging from 100 Hz to 1 500 Hz is within 1.2 dB. The experimental results confirm that the novel Φ-OTDR technology can enable quantitative measurements of multiple acoustic information in real time.

In order to solve the problem of positioning in UAV flight and landing guidance scene, this paper presented a ranging and positioning method based on wireless ultraviolet lights. The method analyzed communication models of wireless ultraviolet Line of Sight(LOS) and Non-Line-of-Sight(NLOS), thus deriving the distance algorithm for LOS and NLOS communications. The location coordinates of unknown nodes were solved by the four node localization algorithm. Using a 255 nm UV LED as the light source, a PMT as the receiving device and a 10 kHz square wave signal as the ranging signal, ranging experiments under different weather conditions were performed. The experimental results indicate that the ranging error is less than 5 m in LOS communication with the ranging distance of 0~100 m. In NLOS communication, the effective ranging distance is reduced to 0~70 m due to the influence of multipath scattering. When the transmitting elevation and receiving elevation is less than 10°, the ranging error is less than 5 m, otherwise the effective distance decreased significantly with the increase of transmitting and receiving elevations. In general, the algorithm can provide navigation data for unmanned aerial vehicles with a high accuracy when the GPS cannot work normally, thus can meet the requirements of autonomous landing and flight guidance.

In order to suppress the speckle effect of target surface induced by laser illumination, and to increase the image quality of laser active imaging, the multibeams illumination technology was adopted and its effect on rejection of returning wave speckle was investigated theoretically and experimentally. The splitting illumination model of multibeams was established. The speckle suppression principle of multibeams illumination in entire speckle superposed field and partly speckle superposed field was analyzed. On the basic of the principle and theory model, the simulation experiment system was set up, where the returning speckle contrast of single beam and multibeams illuminated on three different roughness targets with different splitting ways was evaluated by adjusting the energy distribution. The experiment result indicates that the contrast of homenergic four beam can be reduced to half of that of single beam, which complies to the theory analysis and proof the feasibility of speckle rejection using multibeams illumination.

Under the premise of not increasing size of code, layout form for reading head of angle measurement sensor was researched to develop high-precision angle measurement sensor with miniaturization. Through harmonic analysis for angle measurement error, the suppression principle of layout for multiple reading head to angle measurement error was deducted and analyzed in detail. On the basis of intensive study on various typical layout modes for multiple reading head, a mixed layout method for reading head combined with odd-number head and even-number head was proposed to eliminate more and higher order errors and improve accuracy of angle measurement sensor. The experiment result shows that at the time of adopting layout form for uniform distribution of three, four and six reading heads, angle measurement accuracies of angle measurement sensor are respectively 15.44″, 9.72″ and 8.96″; when adopting optimized layout form of six reading heads, angle measurement accuracy can reach 7.7″. Thus it can be concluded that optimized layout for multiple reading head can effectively restrain angle measurement error and improve measurement accuracy.

In this paper, a kind of Stack Giant Magnetostrictive Actuator (SGMA) was designed, then the magnetic field distribution model of SGMA was established, and analytical research on model was conducted to increase uniformity of bias magnetic field of GMA. Firstly, given the characteristics and drawbacks of infliction way of bias magnetic field in traditional GMA, structural form of alternative arrangement for permanent magnet and GMM rod was adopted to design SGMA. Then by combining magnetic circuit model with Biot-Savart law, magnetic distribution model can accurately describe magnetic feature of SGMA was establish. Subsequently, the influence of different parameters on magnetic field distribution feature was analyzed on established model, and structural design method of SGMA was proposed. Finally, model verification experiments were conducted. The result indicates: maximum relative error is lower than 4% when using established model to describe magnetic field distribution of SGMA; maximum relative error is lower than 5% at the time of predicting output displacement of SGMA. The established model in this paper can describe work state of SGMA accurately, increase system precision of SGMA and provide reference basis for structural design of SGMA.

To improve the theoretical calculation precision of permeability coefficient for porous material, solving model of permeability coefficient was established on the basis of fractal theory, and solving method of fractal dimension was optimized by image processing. Firstly, after photographing surface image of four kinds of porous materials with a scanning electron microscope, fractal dimension was solved by box dimension method and influence of image size and amplification factor of porous material on solving precision of fractal dimension was researched. Secondly, according to Darcy law, fractal theory and corrected Hagen-Poiseulle gas equation, solving equation of permeability coefficient without correction factor was established, resulting in calculation formula of fractal theory for porous material was improved. Finally measuring platform of permeability coefficient was established, and correctness of theoretical derivation was verified on the basis of judgment result of flow state. The result indicates: with increase of image size and decrease of amplification factor, fractal dimension approaches true value increasingly; theoretical calculation value of permeability coefficient is more accurate after theoretical relation without experience parameters of maximum diameter of hole, minimum diameter of hole, tortuosity, fractal dimension and permeability coefficient is established. Error value is 5%～8% through comparing theoretical calculation value of permeability coefficient and experiment value, which can satisfy error requirement of measuring for material permeability coefficient.This paper provides new ways of obtaining material permeability coefficient accurately.

In the measurement practice of gear helix deviation, there is phenomenon that big difference often appears between helix deviations for different gear teeth. In order to improve the measurement accuracy of gear helix deviation, influence laws of installation error of mandrel and gear on gear helix deviation were respectively researched. Firstly, mathematical models of influences of mandrel installation eccentricity and tilt error as well as gear installation eccentricity and deflection error on gear helix deviation were respectively established. Then precision test experiment of gear helix deviation was conducted by using manufactured flat washer (1#, 4#) and tilt washers with tilt errors respectively of 5.5 μm/45 mm (2#) and 11.9 μm/45 mm (3#). The following result was achieved: the difference of maximum for helix slope deviation fHβ and theoretical model was 0.17 μm and relative error was 7% to adopt 2# tilt washer; the difference of maximum for helix slope deviation fHβ and theoretical model was 0.06μm and relative error was 1% to adopt 3# tilt washer; while shape deviation ffβ of gear helix was basically fixed in the two tests. Experimental result indicates that measured result of deflection error of gear installation on helix deviation is basically the same as theoretical value to verify accuracy of established mathematical model. Error compensation method of compensating helix slope deviation difference through adjusting deflection error of gear installation according to established mathematical model is proposed, which is of important research meaning to develop high precision gear artifact.

Increasing applicable scope of Reynolds number and mixing strength is development trend of micro-mixer design. A passive-type micro-mixer of 3D-asymmetrical rhombus was designed and manufactured on the basis of asymmetrical separation and reconstruction mixing principle, and change of mixing strength and mixing state were studied by means of numerical analysis and visualized experiment. Results indicate that in scope of low Re (0.01~10), the mixing between two components is dominated by diffusion mixing, with increase of Re, the influence of flow speed on mixing strength decreases to a certain degree; in higher scope of Re(10~200), influenced by increase of flow speed, the inter-fluid imbalance inertia collision becomes the main factor affecting mixing, and mixture strength gradually increases and tends to be stable with increase of flow speed. Take the micro-mixer with Re scope of (0.01~200) as research object, influence of structural sizes such as width slit ratio W/sS, divided-compound angle θ, width-to-thickness ratio H/S, etc on mixing strength was analyzed. By taking into consideration of fluid mixing strength and change of passage pressure drop comprehensively, the optimum structural size of passage was designed as Ws/S=0.2, θ=45°, H/S=0.5, and mixing strength of micro-mixer can maintain above 78%. Compared with traditional planar symmetrical separation-and-reconstruction mixer, mixing strength of designed passive-type micro-mixer of 3D-asymmetrical rhombus is greatly increased, which verifies the effectiveness of the structure.

For the large aperture reflecting mirror with flexible strip support, static friction between large aperture optical reflecting mirror and support stripe has great influence on surface distortion of reflecting mirror. Moreover, the influence is hard to measure in quantity directly. In consideration of the fact that ambient temperature fluctuation would lead to state change of static friction, relational expression of temperature-static friction was established. On the basis of the relational expression, the influence of static friction on reflecting mirror surface distortion could be deduced though measuring reflecting mirror surface distortion at different temperatures. By taking a 1.2 m SiC light-weight reflecting mirror with flexible strip support mechanism as research object, the surface distortion measurement of flexible strip support mechanism was conducted at different temperatures with a interferometer, and corresponding coefficients of temperature-static friction relation were calculated with measured data; simulation analysis on force condition of stripe support mechanism was conducted with ANSYS software. The results indicate that measured result and simulation analysis result have good consistency, thus influence of static friction on mirror distortion of large-aperture SiC light-weight reflecting mirror can be deducted accurately in this proposed method.

Due to the influence of vibration caused by magnetic field of excitation coil, the strain measured by test device of Fe-Ga magnetic characteristics is usually larger than actual one. This paper explained and verified the above phenomenon. By analysis the measurement process of dynamic magnetostrictive strain of Fe-Ga alloy, the original testing device was improved by fixing one end of the sample on pole head and adjusting excitation magnetic field near saturated magnetic field of the sample, resulting in error caused by mechanical vibration on the test of dynamic magnetostriction was eliminated. Static and dynamic magnetostriction test experiment were conducted by adopting multi-parameter magnetic test system and original and improved test device of Fe-Ga magnetic characteristics. Experimental result indicates that the improved test device of Fe-Ga magnetic characteristics has the ability to accurately measure the dynamic magnetostriction of sample with low saturated magnetic field. Dynamic magnetostriction of Fe-Ga at the effect of bias magnetic field of 2.7 kA/m was also tested in the experiment, and the result indicates: (1) the frequency of strain is the same as that of magnetic field at the effect of bias magnetic field; (2) the hysteresis of strain on the field increases with increase of field frequency; (3) λ～H curve is of elliptical and the area increases with increase of the frequency. The proposed improved device can effectively eliminate the strain caused by vibration.

As the carbon-fiber reinforced polymer (CFRP) composite material common used for truss structure of space camera has the drawback of low thermal conductivity, great temperature gradient is easily formed in the truss. The method that copper grid was adhered and aluminum film was pasted on the surface of CFRP was proposed to enhance thermal conductivity for carbon-fiber truss of a space camera. Firstly, thermal conductivity model of CFRP was established. Then heat transfer models for four kinds of plates(including bare CFRP plate, two plates respectively pasted aluminum films of 0.05 mm and 0.5 mm, and one plate adhered copper grid with equivalent diameter as 0.08 mm on the surface) were established, and corresponding thermal analysis was conducted on the four different states of plates. Then equivalent thermal conductivities of four kinds of plates were obtained through finite element software. Test result indicates: equivalent thermal conductivity could be improved in different degrees through adhering copper grid/pasting aluminum film on the surface of CFRP. Structural equivalent thermal conductivity was increased to 41.3 W/(m·K) to paste aluminum film of 0.5 mm on the plate surface of CFRP. Finally, thermal design was conducted on carbon-fiber truss of a certain space camera according to structural equivalent thermal conductivity in the test. Thermal analysis result of truss structure indicates that axial temperature difference of a single truss rod reduces from 6.8 ℃ to 0.8 ℃, and the temperature uniformity of truss structure is significantly improved.

A kind of magnetorheological (MR) clutch in compression-shear mode was designed and an experiment device was fabricated to test its transmission performance. Firstly, working principle of MR clutch was introduced. Secondly, distribution characteristic of magnetic flux intensity of magnetic circuit was analyzed with ANSYS finite elements analysis software. Finally, experiment testing device was established to test static transmission performance and dynamic response characteristics of MR clutch. Experiment result indicates that influence of rotational speed on torque of MR clutch is not obvious, while influence of current and normal stress on torque of MR clutch is great. Moreover, the torque increases with increase of current and normal stress; torque of MR clutch in compression-shear mode can reach 146 Nm, increasing by about 6.6 times than that in only model of shearing when mormal stress is 150 kPa under current of 1.0 A and rotational speed of 40r/min; response time constant decreases with increase of the current (current is less than 0.6 A), later current influence is not obvious; response time decreases with increase of normal stress and rotational speed; overall connection response time is within 77 ms. Transmission performance of the developed MR clutch in compression-shear mode is well and control is sensitive.

The precision hot embossing of microlens array needs heat preservation time, leading to poor efficiency. Hence, a rapid hot embossing of micro-arch array on polymer light guide plate (LGP) is proposed using a heated microgrooved die core. It only performs a thermal deformation for micro-scale surface layer flowing inside microgroove spaces without changing macro workpiece sizes. The objective is to develop the hot embossing process of microlens array of micro-optics with high machining efficiency and low energy-consumption. First, the micro-optics performances of LGP are analyzed with regard to the shape and size of micro-arch array; then the micro-grinding is employed to machine precision and smooth microgroove array on steel die core; finally, the rapid and accurate forming process of micro-arch array and its micro-optics application are studied. The micro-optics analyses show that the height and distribution density of micro-arch array greatly influence the micro-optics performance. The hot embossing experiments show that by using a microgroove array die core with an average groove depth of 104 μm and an average groove angle of 121°, the 50-μm-height 3D micro-arch array may be precisely formed in 3 s in the case of 12 MPa in loading pressure and 110 ℃ in loading temperature. Compared with the general screen-printing 2D micro-dot array with 8.2-μm-height, the LGP patterned with micro-arch array increases the LGP irradiance intensity by 21% and its uniformity by 27%, respectively. It contributes to the micro-optics precision design and fabrication applied for LED lighting industry.

In order to improve the performance of the aviation photoelectrical stabilized platform, an adaptive robust control (ARC)-based VCM-FSM (voice coil mode-fast steering mirror) was proposed. Firstly, the current loop was adopted into electric circuit to simplify the complex model of motor into a first-order model and guarantee the stablity of output torque of the motor. Secondly, on the basis of conventional PD controller, ARC algorithm was introduced to control the position of VCM-FSM and restrain disturbance. Finally performance tests were conducted in bandwidth test experiment, disturbance restrain experiment, stabilization experiment of visual axis and robust experiment. As comparison, experimental results of DOB-based VCM-FSM and PCT-FSM were offered at the same time. As compared with the DOB-based VCM-FSM, parameter robust characteristic of proposed control method is stronger. Meanwhile, as compared with traditional PZT-FSM, it also exhibits longer travel range, much simpler driving circuit, and lower power consumption while the precision of LOS stability remains almost the same. Moreover, visual axis stabilization precision of ARC-based VSM-FSM within 5 μrad(RMS) can all be achieved under disturbance influence of any frequency within 80 Hz. This outstanding performance can be maintained at the temperature of -40 ℃ to 50 ℃, which is superior to the DOB-based VCM-FSM and can fulfill the requirements of a high precision aviation photoelectrical stabilized platform. Thus it has high practical value on improving anti-disturbance performance of control system for optoelectronic stabilizing platform of aviation.

Aiming at the process of low-dose photon counting imaging with Poisson-Gaussian mixed noise, a sparse reconstruction method of integrating data fidelity term and sparse constrait term is proposed. Firstly, based on the hypothesis that Poisson and Gaussian noise are mutually independent, the sparse reconstructing objective function based on integrating data fidelity term and sparsity constraint term is established. Based on patch clustering, the improved greedy algorithm is applied to implement sparse decomposition and dictionary update. Finally, a clean image is obtained by alternating iteration. Contrast experiments on images corrupted with strong Poisson-Gaussian mixed noise show that the average PSNR of image reconstructed by the proposed method increased by 5.5% more than those of the contrast methods, moreover, their MSSIM increased significantly. The experiment results demonstrate that the proposed method has better image restoration and denoising effect for low photon counting image with strong Poisson-Gaussian mixed noise.

A vibration isolator with relaxation type damping was proposed for micro-vibration isolation from satellite remote sensors to get better high frequency attenuation rate as well as the resonance control capacity. A vibration isolation model of the relaxation type viscous damping was established. The absolute transmissibility of the vibration isolator model with relaxation type damping was deduced and compared with that of a traditional Kevin vibration isolation model. The vibration isolator with relaxation type damping was designed by adopting a bellow as spring and an elongated hole to generate damping force. The damping coefficient was solved analytically for the designed structure. The transmissibility experiment for the vibration isolator was conducted. According to the experimental results, it shows that the vibration isolator with relaxation type damping provides a high damping force at the resonance frequency, which controls resonance amplification factor to be with two times. Moreover, it provides a low damping force at the higher frequency, which allows the vibration attenuation at 100 Hz to be over 95%. These results are identical to that of the theoretical prediction and demonstrate that the performance of proposed Vibration isolator is superior to that of traditional ones. The research on vibration isolator with relaxation type damping can be as a good instruction to design and apply such kind of vibration isolator in satellite remoter vibration isolation fields.

A model for evaluation of the capability index of information acquisition in remote sensing was proposed. The main factors affecting the satellite remote sensing information acquisition ability, including orbit, imaging resolution, attitude maneuver ability were analyzed. By taking the comprehensive target identification as the standard, an evaluation model of ability index for the information acquisition which covers imaging resolution, target revisiting and attitude maneuver was established. The implementation process and the specific steps of the model were introduced. The model integrates two dimensions of target acquisition quantity and target acquisition precision, which could evaluate the practical physical significance comprehensively for remote sensing satellite information acquisition capability. With the Worldview series, Pleiades and a number of domestic and foreign optical remote sensing satellites as examples, the validity of the evaluation model was verified. The results show that the index model assess accurately the performance of different remote sensing satellites. It points out that the resolution of WorldView-3 is Less than twice as much as that of the WorldView-1, but its information acquisition capability is 3 times that of the WorldView-1. The resolution of Superview-1 is higher than that of the Pleiades, but its information acquisition capability is less than half of the latter. The validity of the evaluation model was verified, which demonstrates that evaluation model can provide a quantitative reference for the comprehensive evaluation of optical remote sensing satellites, and can provide a support for the evaluation of new optical remote sensing satellites.

To detect ship targets accurately, a new method to detect ship targets on sea surface was proposed based on multi-feature and multi-scale visual saliency. Firstly a scale-adaptive top-hat algorithm was used to suppress the interference of clouds and oil. Then, the double-quaternion images are constructed by using double-color spatial features and edge features to detect the saliency of ships. This method makes full use of the double quaternion images, so it can be operated at the same time in a number of channels, and can save operation time to guarantee the characteristics of different scale characteristics. Furthermore, the method also uses the character that the human eye focused on the different targets for image with different sized in implement of the up-down sampling to avoid the leak overlapping in image detection. When the last saliency map is obtained, the ships were segmented to ensure the target location by using the OTSU algorithm, and then the ship target was marked and extracted in the original image. The experiments were analyzed in the several sea conditions. Experimental results show that the algorithm eliminates the interference of cloud, fog and oil pollution and ship targets are detected accurately. With this algorithm, true rate iss 97.73%, and the false alarm rate as low as 3.37%. Compared to other frequency domain saliency detection algorithms in ship detection, this algorithm has obvious advantages.

As the non-rigid image registration methods can not meet the requirements of registration accuracy and registration time simultaneously, three kinds of improved non-rigid registration methods are proposed based on image characteristics and image gray. These non-rigid registration methods were based on the Circle Descripto increases Feature (CDF), Dynamic Driving Force Demons (DDFD) and image characteristics and optical flow, respectively. In CDF method, feature points were extracted from the images, and the circle descriptor is used in the method instead of square descriptor in classical methods, by which the rotation invariance was maintained and the speed of the registration was increased. In DDFD method, the driving force was changed by introducing the driving force coefficient, so that the registration time and registration accuracy were improved effectively. In registration methods based on image characteristics and optical flow, the feature points were extracted from a float image and a reference image by using registration method based on image characteristics, and these extracted feature points were used to get a coarse registered image (feature level registration); then the optical-flow method was used to register accurately (pixel level registration) for the coarse registered image and to achieves the purpose of taking account of the registration accuracy and registration time. The experiments on checkboard images, natural images, brain MR images and liver CT images were performed and the results show that the proposed methods are better than the classical methods such as Scale-invariant Feature Transform (SIFT), Speeded-Up Robust Features（SURF）, Demons, Active Demons and Total Variation Regularization/L1 norm (TV-L1) in registration time, registration accuracy and adaptability for large-deformation images.

Calibration methods of traditional star sensors ignore the additional errors from the interaction between optical parameters and distortion coefficients. This paper proposes an autonomous calibration algorithm based on nonlinear optimization to overcome the problems mentioned above. Firstly, the algorithm ignores the distortion to construct a target function, and the Levenberg-Marquardt nonlinear optimization algorithm is used to optimize the optical parameters of the star sensor. Then, the optimized optical parameter estimation iss used as the ideal value, and the lens distortion coefficient of the camera is estimated by the linear least square method. Finally, the parameters obtained by the first two steps are used as initial values to construct the target function, and the optical parameters and distortion coefficients are optimized by using Levenberg-Marquardt algorithm. Simulation and comparison experiments are performed in combination with least square method and Samman method, and results show that the maximum residual obtained by the algorithm is 0.015 pixels under the same test condition, and the accuracy is higher two orders of magnitude than that of the other two calibration methods. Moreover, the field experiments show that the proposed method effectively improves the performance of star sensors.

An improved regularization blind restoration method based on L0 sparse prior was proposed to overcome the image blue from camera shake. A new optimization mode on the basis of inherent property which the gradient distribution of the blurred image is denser than that of the clear image and the sparse of the dark channel is relatively smaller. Aiming at the highly non-convex of L0 norm and nonlinear minimization problem in the dark channel sparse optimization process, an approximate linear map matrix based on look-up tables was proposed, and the linearized L0 minimization problem was solved by half-quadratic splitting methods. Finally, the fast Fourier transform was used to do iterative operation alternately for the fuzzy kernel and the clear image in frequency domain to obtain the restored image. Through experiments on several different types of blurred images, the results show that average gray level gradient is up to 11.411, the image entropy is up to 7.304, and it only takes 807s to process 365×285 images. The improved regularization algorithm effectively suppresses the ringing effect near the edge of the image, retains the integrity of clear details, improves the speed of operation significantly. The algorithm is suitable for all kinds of image restoration.

As collaborative trackers based on traditional fusion strategy has poor robustness in complex environments, a novel adaptive interactive fusion tracking strategy based on the online updated transition probability matrix in a multiple model particle filter framework was proposed. Firstly, an iterative updating equation was obtained based on minimum mean square error estimation method based on the Bayes theory. Then, the numerical solution of the iterative equation was obtained by numerical integration algorithm. Finally, with the updated TPM and re-sampling technology, the adaptive interaction of prior state distributions for different trackers was achieved to guarantee the target state of transmitted particles with larger weights. Tracking experiments were performed in complex environments. The results demonstrate that the proposed adaptive interactive fusion strategy improves the correction function for Particle prior state and effectively avoids the ‘tracking drifting’ problem from error accumulation. So, the robustness of proposed collaborative tracker is more better than those single trackers or collaborative trackers based other fusion strategy.

This paper focuses on the accumulated positioning errors and poorer fault tolerance of Inertial Navigation System(INS)in an underwater vehicle. The phase difference position of the underwater acoustic Ultra Short Base Line(USBL) was analyzed and the solution process and coordinate transformation of INS / USBL navigation was deduced based on real time positioning and attitude error angle information obtained by the INS. Finally, in combination of a INS / Doppler Velocity Log （INS/DVL）filter, the applications of the INS/USBL/DVL integrated navigation federation filtering in three kinds of information fusion algorithms. The fusion algorithm was verified by MATLAB simulation. The results demonstrate that algorithm is able to inhibit filtering divergence of INS. It makes the best of the parameter information obtained from three navigation system, and shows low state dimension and fast convergence speed. Among them, the navigation system based on precision factor has minimum system error. The fault tolerance verifies that the INS/USBL/DVL integrated navigation system still can provide effective navigation parameters at a higher navigation speed when the UBSL was broken. It concludes that the proposed INS/USBL/DVL integrated navigation federation filtering algorithm is able to provide navigation parameter information with high fault tolerance and stability for underwater vehicles effectively.

Ormia ochracea with an extremely short baseline is able to locate its specific parasitoid host (cricket) precisely. On the fact mentioned above, this paper establishes a biomimetic coupled processing system based on the Ormia ochracea to improve the performance of miniature direction finding (DF) devices. According to the mechanical model of coupled auditory structure of the Ormia ochracea, the time domain response and frequency domain response of the auditory structure were deduced and their phase difference responses were computed. Then, a signal processing model for biomimetic coupled processing system was established, and its performance was analyzed. The results show that proposed system can be used to handle arbitrary frequency signals and enhance the phase difference between signals by setting proper coupled parameters, particularly, the phase difference can be amplified about 18 times under ideal conditions. For certain coupled parameters, there exists the most sensitive frequency and the most sensitive incident angle in the system, namely, the most sensitive frequency makes the phase difference between responses arrive to 180°. Moreover, the gradient of phase difference between responses reaches its peak at the most sensitive incident angle. The biomimetic coupled processing system may adopt two kinds of mechanisms to measure directions: one is direct DF with the measurable phase differences, and the other is indirect DF by baseline rotating using the most sensitive frequency and the most sensitive incident angle concepts. This paper provides a theoretical foundation for much wider applications of the biomimetic coupled processing systems.

A method was proposed to obtain characteristic information in a welding process by visual sensing and to predict the weld width of weldment bottom surface by using a neural network model. A workpiece made from mild steel SS400 was welded by a high power disk laser. In welding processing, the weld conditions were changed, including laser welding power, welding speed and welding route and two high speed cameras were used to capture images containing characteristic information on both top surface and side surface of weldment simultaneously. In order to get a better characteristics extraction, the colour space of a RGB image was changed into NTSC (National Television Standards Committee) colour space, then both RGB image and YIQ image were separated into their colour components, filtered to denoising and processed in space domain. The weld characteristic information was extracted, including spatter, weld pool and metal vapour and the effect of weld route on characteristic information was researched. Finally, a LMBP (Levenberg-Marquardt Back Propagation) neural network model including three layers and one hidden layer was established. The obtained characteristic information was taken as input, and the weld width of weldment bottom surface was predicted. The results show that when the welding penetration is unstable or lack of penetration, the fitting degree of LMBP neural network is greater than 0.83, the maximum training error mean is 0002 8 mm, and maximum actual error mean is 0.225 6 mm. It concludes that the prediction model has good accuracy and stability.

To accurately reflect the geometric imaging relationship of cameras, a self-calibration method was proposed based on simplified Brown nonlinear camera model and improved BFGS (broyden-fletcher-Shanno) algorithm. In this method, the linear camera model and the distortion model were fitted into a nonlinear model, and the nonlinear model parameters were constrained by fundamental matrices of the linear model to obtain a set of nonlinear constraint equations.Then, based on new quasi-Newtonian equation, an improved BFGS algorithm suitable for nonlinear internal parametric constraint equations were presented and the internal parameters of the equation were solved. By using the proposed model and algorithm, the calibration method improves the accuracy and robustness of the calibration results in fewer iteration times and noise conditions. The convergence analysis and robust analysis in with or without noises show that the reprojection error is less than 0.4 pixel when the noise is not greater than ±3 pixel. A real image experiment was performed by calibrating camera parameters and calculating the projection error, and the results show that the calibration precision error is less than 0.06%, and the re-projection error is 0.35 pixel, which verifies the effectiveness of the proposed method. It concluds that the method is applicable to image processing, mode classification and scene analysis in computer vision field.