In order to improve the domestic roll-to-roll manufacturing level of optical microstructure films and change the fact that the manufacture of nearly all ultra-precision roll machining equipment is monopolized by foreign companies, the first domestic ultra-precision roll lathe is designed and integrated successfully with process experiments for typical microstructures. First, according to the characteristics of optical microstructures for roller mold processing, the key technologies of machine tools were analyzed. From this analysis, the machines structure and motion control system are designed. After integration of the machine tool system, debugging of the motion control system and linear axis motion resolution tests are carried out. Second, the straightness and positioning accuracy of each axis are measured and compensated using a collimator and laser interferometer. Finally, parameter optimization tests for reducing the burr height when manufacturing V-groove arrays are performed through a cutting experiment by using a multi-step cutting method. Experimental results indicate that the linear motion resolution of the ultra-precision machining lathe can reach 5 nm, full-stroke bi-directional positioning accuracy of the linear axis is better than 1 μm, and the V-groove micro-structure array processed with a final cutting depth of 1—2 μm has excellent surface quality.

In this paper, we present a general review of the development of a high-power all-fiber side-pumping combiner developed over the past few years. Our main focus is on the fundamental mechanism of operation, in addition to a review of the development and the challenges associated with the realization of a functional side-pumping combiner, which was fabricated using a tapered-fused technique. This technique can potentially be exploited as a side-pumping method because of its high coupling efficiency, high signal transfer efficiency, and its capability of handling high pump powers up to the kilowatt range. This review includes an overview of theoretical and experimental results, the challenges associated with fabrication, an investigation of the loss mechanism, and an overview of the performance improvement of the combiner. In summary, a new (2+1) ×1 side-pumping combiner for a 2.5 kW fiber laser based on tandem pumping is introduced. The presented results suggest that the side pumping combiner has a high power loading capability as well as a high coupling efficiency when applied in a tandem-pumping scheme, due to the high brightness of the fiber laser as the pump light source.

An interferometer is proposed for high precision straightness measurement. This instrument consists of a Koester prism, corner cube, quarter-wave plate, wedge prism, and wedge reflector. The wedge prism acts as the straightness sensor while the corner cube and wedge reflector are used to reflect the measurement signal. During operation, the two frequency interferometer signal is decomposed into four beams with a spacially symmetrical measurement structure. The beams have common paths, which result in an improved stability of the interferometer and the minimization of dead path. A measurement resolution of 17.71 nm was determined with a phasemeter which has a resolution of 2π/512 and a wedge prism with one wedge angle. The setup does not require a mirror, which typically has a size that is comparable to that of the stroke of a linear moving stage. However, it can be used to measure straightness with very high resolution. The theoretical and experimental results demonstrate that the crosstalk of the Abbe error caused by pitch, yaw, and roll angle is avoided due to the spacially symmetrical measurement structure. The proposed interferometer has several advantages such as fewer optical components, a simple structure, ease of use, and repeatability of measurement results.

Turbine blade tip clearance varies in a three-dimensional (3-D) direction in an aero-engine. The traditional fiber-optic sensor can only measure the radial clearance, and the results are affected by the interdimensional coupling effect. In this paper, a sensor array consisting of three of the two-circle coaxial fiber bundles, arranged along the isosceles triangle layout, was used to realize the decoupling of blade tip clearance 3-D parameters. These parameters included radial clearance, axial deviation angle, and circumferential deviation angle from the sensor outputs, leveraging a BP neural network method. A special fiber-optic sensor array and signal conditioning circuit were developed, and the experimental tests for static and dynamic properties of the measurement system were performed. Experimental results indicate that the maximum error of radial clearance static measurement is 47 μm with a standard deviation of 10 μm. The maximum error of axial and circumferential deviation angle static measurements are 0.49° and 2.32° with standard deviations of 0.13° and 0.36°, respectively; the measurement system performed well in terms of repeatability and reliability. The dynamic measurement standard deviation of radial clearance is less than 18 μm with axial and circumferential deviation angles of less than 0.2° and 0.5°, respectively. This satisfies the demand of real-time detection of blade tip clearance 3-D parameters of turbine blades of an aero-engine.

Because the fluorescence signal induced by a cancer cell is weak and has a wide spectral range, along with the detection limits of existing microscopes, this paper proposes fluorescence microscope objectives with a spectral range of 450—800 nm and a numerical aperture of 0.95. The objective used “+ + -” structure because the correction of aberration, caused by a wide spectrum and high numerical aperture, is difficult. The objective must include a set number of lenses and the alignment process is difficult; the front group, considering more than 90% focal power, is set as the sensitive component, which is useful for adjusting the alignment process; the middle group is the weak focal component, which is used to correct the secondary spectrum caused by the large numerical aperture, so that the difficulty in manufacturing the apochromatic elements is reduced significantly. The posterior group is the negative component, which is useful for flattening the image field and increasing the working distance of the objective. The objective parameters are as follow: the total length is 58 mm, the working distance is 0.21 mm, the magnification is 40×, and the full field is 0625 mm. The results show that the image quality nearly reaches the diffraction limit and the distortion is less than 0.2%, meeting the surveilling requirements of the weak fluorescence signal from cancer cells.

In order to prepare well-ordered micron and submicron grating fold structures for simpler stress measurement, tunable grating was prepared by a self-assembly process based on a rigid film/flexible substrate. A polydimethylsiloxane (PDMS) substrate was prepared on polyethylene (PET) by the spin-coating method; then, the curved PDMS-PET substrates were processed in oxygen plasma. After appropriate surface treatment time in plasma, the curved substrates were flattened, and well-ordered wrinkling shape gratings were obtained due to mechanical buckling instability. Then, according to the principle of grating spectral, proper stress measurement can be achieved when a tunable grating structure was used for stress testing. The experimental results show that: When the curvature radius of the grating is 1.4 mm, the wavelength of the tunable grating folds prepared from 0 to 10% of the strain ranges from 452 to 507 nm; when the radius of curvature of the grating is 5.6 mm, the wavelength of the tunable grating folds prepared from 0 to 15% of the strain ranges from 498 to 572 nm. The proposed method of tunable grating fabrication is a low cost, simple process and enables mass production technology. It is a potential method to prepare variable spacing grating. Its expected applications in future include spectrometry, optical communication, and other related fields.

To address the problem of heat accumulation of optical crystals in solid-state lasers and ensure the stability of its work performance, a temperature-control method with high temperature control accuracy and large temperature control range, based on a semiconductor Thermoelectric Cooler (TEC), was proposed. Based on the operating characteristics of the TEC, the heat effect and the heat transfer mechanism of the optical crystal, a temperature-current double closed-loop temperature control model was established. The proposed model controls both the electrical characteristic variations of the working TEC and the temperature variations of the optical crystal. A high-precision and wide-range optical crystal thermostat control system was designed and implemented. Experimental results demonstrate that the overall temperature control accuracy of the proposed system is better than ±0.002 ℃, and the system can quickly stabilize at any set point temperature in a wide temperature range of -15—120 ℃, when the set temperature and the actual temperature deviate 20 ℃, a stable convergence time of less than 60 s, which meets the requirements of solid-state laser optical crystal for temperature range and accuracy.

A mirror deformation detection system based on STM32 and using the capacitance detection chip Pcap01-AD was designed to solve the problems of small deformation, high change frequency, and difficulty of accurately measuring the microdeformation that were associated with a large aperture adaptive sub-mirror. First, according to the characteristics of the deformable mirror, which was driven by the voice coil motor, a system scheme for measuring the microdeformation of the mirror using the capacitance displacement sensor was proposed. Then, the hardware and software are designed for the measurement system. The hardware was constructed with the microcapacitance detection chip Pcap01, STM32F103 microcontroller (MCU) system, and power supply circuit. The software part included programming the lower computer in C language to realize digital signal acquisition of the capacitance, communication between the Pcap01-AD and MCU, data conversion of the upper computer and data processing, real-time display, and data storage. According to the structural characteristics of the large aperture telescope sub-mirror system, the experimental platform was designed to enable the conduction of several tests. The test results show that in the range of 50 μm, the measurement sensitivity is 200 pF/3 μm, and the system has high accuracy, small error, and high detection efficiency. Thus, this system can be used in adaptive mirror surface deformation detection and other small-displacement measurements.

The absorption and scattering of underwater light by water are the two main factors influencing image degradation. With a focus on underwater target imaging, this paper proposes an underwater target polarization recovery method by global parameter estimation, using a simplified underwater target polarization reconstruction model, which recovers an underwater target image by automatically estimating global optimal polarization information reconstruction parameters to reduce the impact of the medium on image quality. First, a polarization median filter method and the bright primary principle method were used to estimate the underwater background light polarization degree information and the infinite underwater background light intensity values, respectively. Based on the principle of minimum mutual information, the background light polarization degree information was optimized and estimated. Then, adopting the underwater target polarization image enhancement algorithms to enhance the details of the underwater reconstruction target, a clear figure of the underwater targets radiation information was produced. The experimental results show that the value of the measure of enhancement (EME) in the suggested method increased by 120% on average, compared with the original strength of the figure and that of other underwater recovery methods for the calculation of the performance evaluation index. Further, the image quality was significantly improved. The method has solved the inaccuracy problem while estimating parameters via artificial view and improved the contrast of the restored target image. The method can be used to detect and identify turbid underwater targets.

The quality of optical machine installation in a 1 000 mm aperture reflector directly influences the imaging characteristics of the optical system of a telescope. In order to improve the imaging quality of a large-aperture photoelectric tracking system, the installation and adjustment technology of the reflector was investigated. First, the errors that affect the precision of the principal mirror were distributed. Second, a typical method was adopted to perform a detailed installation and adjustment investigation of the principal mirror with an aperture of 800 mm, and the reflection mirror shape accuracy was measured in real-time, in combination with the specific form of the reflector support structure. Finally, this paper included a comprehensive analysis of the source of installation and adjustment errors and proposed a high-precision method that integrates processing, testing, and installation. The method adopted resulted in an RMS value of λ/40 for the modified profile precision waveform aberration after installation, and significantly improved the efficiency and quality of the installation and adjustment.

The need to improve the accuracy of Line of Sight (LOS) for long-distance, long-focus, and high-resolution airborne photoelectric payloads is crucial. Starting from the basic principles of current, speed, and position three-loop control, our study analyzed torque control of motors, disturbance suppression, and robustness of time delay. First, the drive and control methods of a new brushless motor at the current loop were analyzed, and the corresponding advantages and disadvantages were presented. Second, the problem of active disturbance rejection internal loop and speed feedback external loop control was discussed, and the control performance of visual axis motion speed was improved from a control algorithm perspective. Moreover, considering the constraint of mechanical rigidity on control gain, the corresponding control method was analyzed from the perspective of two-stage control based on fast steering mirrors. Third, the typical position loop control problem of target tracking and tracking control was analyzed, and the corresponding theoretical research progress was discussed from the viewpoint of time-delay feedback control. Finally, some research suggestions on airborne photoelectric payload LOS control technology were presented, combined with the developing trends of current technologies.

Ultra-white glass is a kind of transparent low-iron glass. Owing to its excellent physical and optical properties, it is widely used in many fields such as precision electronics, luxury automobiles, and solar photovoltaic power generation. Because of its hard and brittle characteristics, the outlets of glass micro-holes are easily damaged. In order to improve the machining localization of glass micro-holes and reduce the possibility of micro-hole outlet damage, the micro electrochemical discharge drilling process was studied and optimized. Firstly, according to the principle of electrochemical discharge, the film forming mechanism and the material removal mechanism were discussed, the influence of discharge energy on the micro-holes machining technology of glass was analyzed, and an energy control model of electrochemical discharge was established. Secondly, according to the experiment, the influence of pulse voltage, frequency, and feed rate on the inlet diameter and outlet quality of the micro-holes was analyzed. Finally, the optimized parameters were chosen to fabricate a 3×3 high quality glass micro holes array with an inlet diameter of 172 μm and an outlet diameter of 167 μm on an ultra-white glass workpiece with a thickness of 300 μm; the outlets exhibited no signs of damage. The experimental results show that under the control of the discharge energy, micro electrochemical discharge drilling technology has great potential in glass micro-hole machining.

Cell culture is the basis of cell research. In order to provide an approximate microenvironment for cell culture, a micro-device for three-dimensional and dynamic cell culture was designed. First, a micro-channel network was designed for transportation of liquid. The cell culture chamber was symmetrically arranged in the micro-channel network. A series of “multiple input and multiple output” micro-channels were connected with the inlet and outlet of the cell culture chamber. Second, the physical fields of laminar and porous media in the COMSOL software were coupled, which was used to simulate the velocity of the liquid in the cell culture chamber of the micro-device. The network structure of micro-channels was optimized by comparing the homogeneity and stability of the flow field. Third, the electrohydrodynamic direct-writing technology was used to integrate polycaprolactone (PCL) three-dimensional scaffolds in the cell culture chamber, which constructed the three-dimensional culture space. Finally, after the welding of the micro-device, the fluid flow condition in the cell culture chamber of the micro-device was tested, and cell experiments were performed. The results show that the fluid stability and homogeneity in the cell culture chamber of a “2×2” micro-device are appropriate. The fiber spacing of the three-dimensional scaffolds is 400 μm, the diameter is 80 μm, and the porosity is 64%. Besides, the cell survival rate is more than 90%. The micro-device, which can be used for three-dimensional dynamic cell culture, precisely simulates the microenvironment needed for the survival of cells in vivo. Cells in the culturing chamber grew well; therefore, the micro-device can satisfy the design requirements.

According to the requirements of actuators for novel electro-hydraulic servo valves (EHSVs), this paper proposes a design for a special stack giant magnetostrictive actuator (SGMA). In order to compensate for the nonlinear property of the SGMA, a controlling strategy was proposed and verified by simulation and experimentation. First, with permanent magnets (PMs) and short giant magnetostrictive material (GMM) rods located iteratively, a highly uniform bias magnetic field was obtained in the SGMA. Then, based on the structure of the SGMA, a multi-DOF model was established to describe the dynamic properties of this actuator. In addition, a control methodology was developed, which combines model predictive control and sliding mode control. Finally, to validate the proposed controller, both simulation and experimentation are conducted, and the results indicate that the proposed controller can realize the ultra-precise control of the SGMA. In the step control experiment, the system achieves stability within 1.5 ms with no overshoot or steady-state error. In the sinusoidal control experiment, the maximum tracking error of the system is approximately 0.83 μm, 6.9% of the total output of the SGMA, proving that the model predictive sliding mode control can significantly reduce the nonlinearity of the SGMA.

A biaxial bipod flexible support structure was designed based on flexibility analysis parameter optimization to improve the accuracy of a mirror surface subject to ambient temperature fluctuations. First, the flexibility of the mirror support structure was calculated and analyzed, and a flexibility formula for the leg-and-mirror assembly was deduced. Then, to ensure the axial support stiffness and unloading ability of the mirror, a set of flexible support structure size parameters were calculated with a diameter of 200 mm. Finally, the flexibility formula, dynamic characteristics, and temperature adaptability of the support structure were analyzed and verified by finite element analysis and vibration tests. The results show that the error between the theoretical and finite element analysis values is less than 10%, under a certain force. The first-order frequency of the component, obtained by the vibration test, is 358.5 Hz, and the relative error of the theoretical calculation is 89%. At a temperature difference of 20 ℃, the value of the mirror surface accuracy is 0.8. The validity of the theoretical model was verified, thus proving that the bipod flexible support structure can reduce the influence of temperature fluctuation on a mirror surface.

A new type of hyper-redundant metamorphic parallel manipulator arm was proposed. It was based on the 3-PUPS parallel mechanism and it could change its configuration by locking driving joints. Error modeling and analysis of the manipulator were done, and the positioning error of the manipulators experimental prototype was measured by a calibration system. First, the design idea of manipulator metamorphosis by locking driving joints was proposed so it could change its configuration and performance according to the requirements of the task. Then, using the closed loop vector loop method with error sources, the manipulators error vector model was established, and the influence of each error source on the moving platforms output errors was analyzed. In addition, according to the influence of each error source on the output errors, the machining accuracy, grade, and tolerances of the manipulators parts were determined. Based on these, the manipulators experimental prototype was developed. Finally, the experimental prototypes errors were measured by a high precision calibration system for industrial robots. The experimental results show: the position errors of the manipulators moving platform were between 0.005 mm and 0.003 8 mm, and the attitude errors were between 0.010° and 0.044°. The position errors were slightly higher than the position repeatability (0.05 mm) of the general industrial robot, and the attitude errors were equivalent to the attitude repetition positioning accuracy (0.045°) of the general industrial robot.

In order to study the power generation characteristics of a Dielectric Elastomer Generator (DEG), a Finite Element (FE) model of the DEG under pure shear mode was established in the COMSOL Multiphysics commercial package. The model was based on the Yeoh constitutive relation of hyperelastic materials, coupled with the electrostatic force inside the generator membrane. The capacitance variation of the generator and the effect of power generation were studied. A pure shear tension apparatus was designed such that the DEG can be pre-stretched in the Y direction. Experiments to study the DEG performance under different pre-stretching conditions were carried out, and the capacitance change and power generating effect were obtained. The simulation data and the experimental results were compared and good agreement was found, therefore proving that the FE model is reliable. This study shows that, under the same tensile conditions, pre-stretching in the Y direction increases the initial capacitance and the capacitance changing rate of the DEG membrane. Hence, when the DEG was pre-stretched at λ=1, its rising voltage was 83 V, while at λ=2, its rising voltage was 252 V, thereby improving the power generation performance. The new research method of DEGs presented in this paper provides a new idea for the design of a generator prototype.

Centrifugal melting casting technology has unique advantages in the manufacture of large aspheric mirror blanks. In this paper, the centrifugal casting process and the quality of the mold embryo were studied in depth by using a self-developed shrinkage model experiment machine. In order to obtain the required mirror blank, the processes of preparing the coating, heating, cooling and preserving the heat of the mold were introduced in detail. The influencing factors of shape deviation in the mirror blank production were studied by comparing simulation values with experimental results. The influence of the coefficient of thermal expansion, cooling rate, diameter, and heating temperature on the surface deviation was analyzed, and a mathematical expression relating these parameters to the deviation was derived. The compensation of the deviation was then calculated in two iterations, thus reducing the surface deviation from 84 to 33 μm. The compensation results met the design requirements. Using centrifugal casting technology, the aspheric mirror embryo can be prepared to meet the vertical deviation specifications of its upper surface in the range of 30~40 μm.

In order to improve flywheel reliability, flywheel fault diagnosis technology was studied. A hybrid fault diagnosis method based on a neural network was proposed, which compares the mathematical analysis model with the flywheel fault diagnosis based on intelligent computing. In this method, the difference between the mathematical model and the original system output was used as the first-order residual. Then, the first-order residual and the system measurements were used to train the neural network. Finally, the second-order residual of the mixed model output was used to detect the system fault. This method was validated using the flywheel injection bus voltage and armature current faults. Under the bus voltage fault working conditions, the hybrid model avoided the divergence problem of current estimation because of the analytical model, which reduced the maximum tracking error by 44% compared with a single neural network model. Under the current fault working conditions, the maximum tracking error of the hybrid model was reduced by 90% and the tracking variance was reduced by more than 10 times under different speed conditions compared with two single neural network models. These results illustrate that the hybrid method can effectively solve the problem of inaccurate fault diagnosis due to the existence of modeling errors in the analytical model, as well as the problem of a single neural network model being unable to adapt to fault diagnosis corresponding to new working conditions because of the lack of training data.

Satellite-borne opto-electromechanical instruments preserve both optical systems and mechanism properties; the former has a strict requirement for temperature, and the latter has a complex structure and is constantly moving. This is a challenge to the test validation of the thermal characteristics. A case study of the indirect validation method of the thermal design for the optical communication terminal is presented. The indirect validation method may be utilized to predict the in-orbit temperature through the thermal analytical model, which is corrected using temperature data in the thermal balance test. The thermal balance test and model correction have been made. Very good correlation was demonstrated between the computed results of the proposed model and the measured test data; a deviation of 81% is less than 5 ℃. Thus, the model is appropriate for use in validation of thermal design. Comparing the thermal analytical results to the flight results, the deviation of 81% is less than 4 ℃. This demonstrates the accuracy and the effectiveness of the indirect validation method, and the method meets the application requirements of optical communication terminals on satellites. The proposed indirect validation method greatly reduces the difficulty of thermal testing and is of certain reference value to altitude-varied and high temperature precision opto-electromechanical instruments.

An Active Disturbance Rejection Controller (ADRC) can suppress uncertain disturbances effectively; however, its parameter tuning is difficult. In order to develop an adaptive linear ADRC, the parameter tuning strategy of a linear ADRC was studied. First, an adaptive tuning algorithm based on observational errors was designed for a linear extended state observer. Then, an adaptive tuning algorithm for the parameters of the ADRC linear feedback component was designed. Finally, by using the Lyapunov method, it was proved that the parameters obtained by the above adaptive tuning algorithms ensure that both the observation errors of the extended state observer and the output errors of the controlled system converge to zero. Experimental results indicate that, under low velocity conditions, the parameters of the proposed linear ADRC can be quickly tuned within 0.8 s; the observational errors of the linear extended state observer are less than 2 nm, and the velocity fluctuations of the precision air-bearing platform are within 5%. The proposed adaptive linear ADRC performs online tuning of the controller parameters, and the controller performance is satisfactory.

This study aims to verify the effect of image-distortion correction on the edge-field-of-view scanning direction in the case of zoom-scanning wide-swath imaging and to simplify the scanning mirror control. First, the nonlinear relationship between the scanning speed of a scanning mirror and the imaging angle of view during zoom imaging is analyzed. With respect to the problem of low stability and the difficulty of realizing control of the variable-speed scanning of the mirror, uniform angular-velocity scanning with a scanning mirror combined with a variable-frame-rate zoom-imaging detector is proposed based on the analysis and realization of a variable-frame-rate detector. Based on the experimental system, the relationship between the focal length of the optical system, the real-time frame rate of the camera, and the imaging field angles at the time of the uniform angular velocity scanning are determined in order to achieve variable-frame-rate zoom scanning imaging. Finally, a wide-swath long-wave infrared (IR) zoom experiment system based on mirror-scanning imaging is constructed using the pre-existing continuous zooming long-wave IR optical system, and the system is used for imaging verification. The results show that by employing a variable frame rate combined with zoom imaging, the image distortion of the edge-field-of-view can be effectively suppressed in the scan direction in wide-swath imaging. The variable frame rate imaging method simplifies the scan-mirror scanning in zoom scanning, and it is confirmed that the zoom-scanning method is able to eliminate the image distortion.

Based on the persistence and exponential decay across scales of Tetrolet coefficients, the Tetrolet-domain universal hidden Markov tree structured sparse prior model was established for compressive imaging. In this model, the statistic distribution of Tetrolet coefficients was presented as the prior with the Gaussian-mixture form, and then, the posterior probability density function (PDF) was estimated by using the factor graph method. In order to solve the problem that the messages passing through the loop factor graph cannot reach stable convergence, the Turbo equalization method was exploited to decouple the factor graph into two parts for estimating the states of compressive sampling and the structured sparse model. Then, the exchange of messages was performed mutually in the two sub-graphs until reaching convergence. Finally, the image was estimated based on the minimum mean-squared error criterion. The normalized mean-squared error of reconstructing the testing image with size 128×128 was -20.97 dB, and the run-time was 45.24 s. Experimental results demonstrate that the proposed algorithm outperforms the algorithms based on the wavelet-domain hidden Markov tree model in terms of reconstruction quality and speed.

A novel feature partition matching scheme for two-view Terracotta warrior images was presented to address the problem of high false matching rate and low feature matching efficiency during 3D reconstruction in this paper. The new scheme was as follows: First, the features of the complete Terracotta warriors image were extracted using the learned invariant feature transform (LIFT) method. Second, the position of the dividing line on the head of the image of the warrior was determined by applying the proposed prior knowledge-based feature point distribution curve, and the extracted features were then divided into head and torso features based on the dividing line. Third, the Euclidean distance was used to perform the regional feature matching, and the random sample consensus (RANSAC) algorithm was subsequently used to filter out the mismatched point set from the matched result set. Experimental results show that in the terracotta image feature extraction and matching, the correct matching rate of the new scheme can reach 98%; the correct matching rate is increased by approximately 20% compared with those of the SIFT and SURF methods, and the repeat rate of the feature points is increased by 10% while the iteration time of RANSAC is decreased by 50%. The new scheme also has better robustness when scale, illumination, and angle are changed in the images. Therefore, the proposed scheme can achieve correct matching of the feature points with sufficient accuracy and has applications in the robust 3D reconstruction of the Terracotta warrior images.

The use of the traditional Vertical Line Locus (TVLL) matching method for the multi-view image matching of ground primitives results in the following shortcomings: obtaining one elevation value and the lack of accuracy validation of the matching results. Considering the above shortcomings, this paper proposes a new multi-view VLL matching algorithm for optical aerial images based on the constraint of object space positioning consistency. The proposed algorithm brings about specific improvements to the TVLL method. First, it improves the image space multi-view matching process of the object space ground primitive. Second, it uses the known object space plane coordinates of the ground primitive as the constraint of the object space positioning consistency, and improves the accuracy validation process of the multi-view matching results. Thus, it can obtain verifiable matching results of the ground primitive. The proposed algorithm was compared with the TVLL method using actual aerial images. The experimental results show that the quantitative indexes of reliability, accuracy, and computing efficiency for the proposed method are 26%, 40%, and 4 times higher than those for the TVLL method. Therefore, the proposed method successfully overcomes the drawbacks of the TVLL method, and obtains a more reliable and effective matching of ground primitives.

A 3-CCD microscope vision system is established to realize micro-target pose adjustment in a high-power laser device with high precision. Image-based visual control method is performed, and the image Jacobian matrix is derived online by using the active movement of the micro-target. The PI controller is used to adjust the micro-targets pose so that the image features are close to the objective position. In the experiments, we compare the image-based visual control method with the position-based visual control method, which was proposed in our previous research by using a 3D posture description algorithm to control the micro-target to the desired pose. The precision of target positioning and orientation error for the image-based visual control method are 0.07 μm and 0.02 μrad, respectively, whereas the precision of target positioning and orientation error values for the position-based visual control method are 0.16 μm and 0.07 μrad, respectively. The experimental results show that the image-based visual control method is robust for system errors such as kinematic and visual calibration errors. Thus, it possesses better precision and stability for target pose adjustment.

In order to improve the reliability and safety factor of driver assistance systems, and achieve pedestrian detection with a higher precision, an improved pedestrian detection method based on a tree-structured graphical model of the human body is proposed, and it consists of an offline training part and an online detection part. First, the corresponding parent-child parts are obtained by defining the symbiotic relationship between human parts, and then the K-means algorithm is applied to the location relationship between part pairs to acquire part types via clustering. For the purpose of taking both intra-class tightness and inter-class differences into account, a hybrid particle swarm optimization algorithm is built with a two-phase fitness function via introducing MSE and DBI. It is not only effective in estimating the number of optimal cluster centers, but also in eliminating the effect of random initialization on the clustering accuracy. Then, the part type obtained using the optimized clustering method is considered as the latent variable. The pedestrian detection model is obtained through solving the latent structural SVM problem. Finally, we estimate the position of human parts and the detection bounding box on multiple scales based on solving the state equation via a dynamic programming algorithm, and obtain the final pedestrian detection result through incorporating the idea of non-maximum suppression. Experimental results indicate that the performance of the proposed algorithm is superior to those of five other pedestrian detection algorithms. In particular, on the INRIA and ETH databases, the loss rate of the proposed algorithm decreased by 8.14% and 5.05%, respectively, compared with that of the pose-original method. Experimental results show that the proposed algorithm has good performance and high accuracy and robustness.

In order to eliminate the effect of the misalignment error between the sensor array and the sensor system of the magnetic gradient tensor system on the measurement accuracy, a method of theoretically precise calibration between all magnetic sensors and reference platforms, which involves rotating a circle around an arbitrary axis of the system, was proposed. The linear correction model of the sensor system error was constructed using two nonlinear transformations without any mathematical simplification, and the ideal orthogonal output of the reference platform and each sensor was obtained with only 10 sets of measurement data in the same rotation period. By constructing the rotation matrix of the tri-axis heel, pitch, and azimuth transformations of the magnetic sensor, the misalignment error correction model of the arbitrary spatial orientation of the sensors was obtained, and the rotation angle was estimated by the least-squares method. In addition, only three sets of measurement data in the same rotation period was necessary for the alignment of the tensor system. The simulation and experiment show that the accuracy of the simulation parameters estimation was close to 100% in the ideal condition. After the calibration experiment, the output of the sensor showed a high overlapping and coaxiality performance, and the RMSE (root mean square error) of the tensor components was less than 30 nT/m. It is possible to improve the measurement accuracy of the differential magnetic gradient tensor system efficiently with simpler steps and less sampling data.

In recent years, several supervised learning methods have been introduced in hyperspectral image (HSI) classification. However, these methods use only spectral information without taking into account the spatial features and manifold structures of HSIs. To overcome this problem, a new classification method was proposed for HSI classification, combining spatial-spectral features and manifold reconstruction. Based on the spatial consistency of ground objects distribution in HSIs, the proposed algorithm used a small number of labeled samples and large number of unlabeled spatial neighbor samples to perform semisupervised learning, and utilized the reconstruction error of test samples in each submanifold to represent the similarities for discriminant classification. Experimental results obtained from the Indian Pines and University of Pavia data set reveal that the proposed method exhibits a higher classification accuracy compared to other classification algorithms under various training conditions, the highest overall accuracy achieved in the two cases being 95.67% and 91.92%, respectively. The proposed method integrates spatial-spectral information to represent the submanifold structure of different land objects, exhibits superior discrimination performance, especially for a small number of training samples, and effectively improves the performance of HSI classification.