Optical fiber hydrogen sensing with high sensitivity and fast response is the future developmental direction in hydrogen sensing technology that will be of great importance in ensuring the safety of hydrogen energy systems. To solve the difficulty in coupling palladium-based hydrogen sensitive materials at the nanoscale to optical devices, Au-Pd core-shell nanoparticle films with rapid hydrogen response characteristics were prepared by aqueous phase synthesis and centrifugal deposition. With an array of Au-Pd core-shell nanoparticle films, a transmission sensing system was established in which optical signals could be coupled with the multilayer nanoparticle membrane array. Consequently, the sensitivity of the sensing system was enhanced by improving the modulation capacity of sensing materials to the optical signal. The experimental results show that the Au-Pd core-shell nanoparticle membrane prepared in this study has a particle size of 48 nm and a Pd layer thickness of 4 nm. The response time of the sensitive film to 4% hydrogen concentration is less than 3 s and exhibited good repeatability and stability in the cyclic test. The sensors sensitivity is increased to a factor of 2.7 higher than that of the single film without affecting its response speed. This research can provide important guidance for the development of high-performance optical fiber hydrogen sensors.

To improve the angular accuracy of the polarization imaging orientation sensor, an error propagation model of the polarization imaging orientation sensor based on Stokes vector was established. The principal point deviation, lens distortion, and CMOS planar inclination were studied. First, the polarization azimuth calculation method was derived according to the polarization measurement principle of the Stokes vector. Then, the effects of principal point deviation, lens distortion, CMOS plane tilt, and gray response inconsistency were explored. The Monte Carlo method was used to analyze the measurement error of the polarization azimuth. The rationality of the model established was verified by simulation and skylight imaging experiments. The experimental results show that the CMOS gray response inconsistency has the greatest impact on the measurement error of orientation. In the range of (-6, 6) pixels, the measurement error is within 0.11°, and the mean has an overall offset of 0.12°. The principle point deviation changes (-1.5, +1.5) pixels, the distortion shifts (-1, +1) pixel, the CMOS plane tilt is (-0.5°, +0.5°), and the CMOS gray response deviation is (-6, +6) pixels; under the joint action, the measurement error of orientation is 0.236 3°(σ). The error propagation model may reflect how the error source impacts the accuracy of orientation measurement, and it provides a basis and reference for optimal design of the polarization imaging orientation sensor.

Differential absorption lidar combined with near-surface ozone monitoring were used to detect the vertical distribution of ozone in Tianjin from Jun. 23 to Sep. 28, 2018. The monitoring results indicated that the ozone concentration at 300 m above ground showed the same variation tendency as the monitoring values obtained near the ground. In the boundary layer of about 300 m to 1 000 m, the ozone concentration increased with height, but it then decreased at higher altitude. The maximum concentration occurred at approximately 1 000 m above ground. The times when the maximum and minimum concentrations occurred in the diurnal variation curve of ozone were delayed with increasing altitude. This phenomenon was influenced by the gradual transmission process of ozone precursors from the near-surface level to the upper air and the wastage of nitrogen oxide attributable to its oxidation reaction during the transmission process. Above 1 500 m, the diurnal variation curve of ozone showed a double-peak distribution. During the episodes of ozone pollution, the pollution zone with high ozone concentration could be detected frequently in the upper air. The ozone concentration at 1 000 m could reach about 570 μg/m3, and the thickness of the pollution zone could be greater than 1 km. The ozone at high altitude could last for days and sometimes did not dissipate completely at night. In this research, a mixing phenomenon of ozone pollution between high altitude and the near-surface level, which increased the degree of ozone contamination at the near-surface level, was identified 23 times during the ozone pollution days.

In a common method for precision detecton of photoelectric encoders, a polyhedron is joined with a photoelectric encoder and light from an autocollimation setup is reflected by the polyhedron; the precision of photoelectric encoder is then measured by reading the numerical value of the autocollimation. A feasible method to improve measurement efficiency was investigated with respect to the main factors that affect photoelectric encoder detection. Firstly, according to the principle of autocollimator detection, a theoretical analysis was conducted by analyzing polyhedron pyramidal error as it effects the reading of the numerical value of autocollimation. Secondly, according to the principle of installation eccentricity of a polyhedron pyramidal, the impact factors that affect the measurement precision of a photoelectric encoder were analyzed in detail. Pyramidal error and installation eccentricity are the main external factors, which are determined by the numerical value reading of autocollimation. A 21-bit photoelectric absolute encoder was detected using a 23-sided polyhedron; the numerical value of autocollimation is read along the y-axis, the detection error of polyhedron pyramidal error is 7.9″, and the detection of installation eccentricity is 0.8″. The testing results indicate the effects of pyramidal error and installation eccentricity on the precision are based on the reading of the numerical value of autocollimation, which offers an guidance to improve detection efficiency.

To develop a weak-light detector that meets the needs of space gravitational wave detection in the future, the performance detection of weak-light detectors was initially carried out, and the performance parameters, such as responsiveness, response bandwidth and background noise, of the detectors were analyzed, for finding out the method and solution of weak-light detector development that can meet the requirements of future space gravitational wave detection. Firstly, according to the design scheme of the space gravitational wave detection, e.g. Taiji, including laser power, orbit design and star spacing, the parameters of the detector required by Taiji were calculated. Next, three kinds of weak-light detectors were developed by the China Electronics Technology Group Corporation No.44 Research Institute, Southwest Institute of Technical Physics, Shanghai Institute of Technical Physics of the Chinese Academy of Sciences with joint efforts of our research group. Finally, the responsivity, response bandwidth, and background noise of the detectors were tested by the low-noise heterodyne laser interferometry system developed by our research group, and the factors affecting the performance of the detectors were analyzed. The experimental results show that the responsiveness of two detectors is better than 1.8×105 V/W and that the response bandwidth is greater than 10 MHz. The ground noise of the three detectors is lower than 10 pm/Hz@10 mHz (0.1 mHz-1 Hz), and the signal-to-noise ratio is higher than 20 dB. According to the aforementioned experimental results, the three detectors have the potential to meet the requirements of the Taiji Pathfinder in terms of responsiveness, response bandwidth, and signal-to-noise ratio.

In practical industrial applications, ambient temperature change is the main source of error in the angular accuracy of the rotating shaft in Portable Articulated Coordinate Measurement Instruments Machines (PACMM). To eliminate this error, this paper proposes a new method to establish a circular grating angle measurement error compensation model with ambient temperature influence factors. First, the harmonic method is used to establish the compensation model of angle measurement error of circular gratings at a specific temperature. Second, the polynomial method is used to establish the functional relationship between the harmonic coefficient and the ambient temperature. Finally, the experimental data at 14 ℃ is used as the verification data that are substituted into both the traditional harmonic error compensation model and the model proposed by this paper. The experimental results show that compared with the traditional harmonic error compensation model, the accuracy of the compensation model proposed by this paper is improved by a factor of approximately four, and the corrected residual peak value is within 2″; this value can effectively compensate the angular error of the circular grating at 10 to 40 ℃.

To measure two Degree-Of-Freedom (DOF) in-plane displacement, a spatially separated heterodyne grating interferometer was designed and built. The optical configuration, measurement principle, and in-plane rotary assembling error of this instrument were investigated. Based on the diffraction of the planar grating and the spatially separated heterodyne grating interferometry, a symmetrical double-diffracted optical configuration was designed and analyzed. The measurement principle and the elimination of periodic nonlinear errors were modeled and studied using Jones matrices for the components. In-plane rotary assembling errors around the z-axis were measured to decouple the measurement results for the x-and y-axes using a two-dimensional rotation matrix. Then, the square and circular paths of the grating were driven to evaluate the functionality of the proposed interferometer. The experimental results indicate that the proposed interferometer is capable of measuring the in-plane displacements of a nanopositioning stage within the range of 30 μm after compensation for the 0.350° rotary assembling error. The error attributed to mechanical vibration is less than 0.15 μm. In this investigation, 2-DOF in-plane measurement was demonstrated using a spatially separated heterodyne grating interferometer.

A novel method of thermal infrared temperature emissivity separation based on correlation and wavelet filtering was proposed to alleviate the ill-conditioned equation problem of thermal infrared temperature and emissivity inversions of hyperspectral data. The basis of the utilized correlation method was the idea that wavelet denoising could be introduced to suppress the error caused by inaccurate atmospheric correction to certain extents, effectively improving the inversion precision of thermal infrared temperature and emissivity. The core goal of the algorithm was to calculate the correlation between the emissivity curves generated by both the atmospheric downward radiation calculation and the wavelet filtering; the temperature with the highest correlation was the inversion temperature. At the same time, correlation was used to calculate the proportion of different scale wavelet signals in the inversion of their emissivity curves. The simulation results show that the correlation wavelet method has an average error of 0.05 K in temperature calculations when the temperature gradient is 0.01 K. In addition, the combined correlation wavelet method is superior to both the correlation and wavelet methods with regards to temperature inversion accuracy and emissivity inversion precision. It is shown that the developed algorithm can restrain the error caused by inaccurate atmospheric correction to a certain extent as well as effectively improve the inversion precision of thermal infrared temperature and emissivity.

To solve the problem that the calibration accuracy of internal parameters of the camera in the large-size single-camera vision system has great influence on the overall measurement accuracy, this paper presents a virtual stereo calibration method for the internal parameters of the camera based on the Mutation Mechanism Particle Swarm Optimization (MMPSO) algorithm. The method is based on a two-stage optimal strategy. Firstly, a camera imaging model is established to estimate the initial values of the external parameters and some internal parameters. Then, the internal parameters are optimized and calibrated by the MMPSO algorithm to determine the final result. To provide accurate calibration control points, a calibration hardware platform was built. An infrared light-emitting diode was fixed on the measuring head of a three-coordinate measuring machine (CMM), which drove the diode to move, and a large-space virtual three-dimensional calibration board was constructed. The experimental results showed that all of the 10 main internal parameters reached the order of magnitude requested by the measurement accuracy, which validated the effectiveness of the method. The results of two calibration methods were measured by equidistant measurement with the single-camera vision coordinate measurement system. The population standard deviation of the three-dimensional calibration method of Janne Heikkila was 0.112 mm, but the population standard deviation of the virtual stereo calibration method based on the MMPSO algorithm was 0.084 mm. The comparison of the standard deviations of the measured data proves that the proposed calibration method is more stable and accurate. This method can meet the requirements of the large-space single-camera vision measurement system for the accuracy of camera parameters, and it has a certain guiding effect on nonlinear optimization problems such as camera calibration in the field of visual coordinate measurement technology.

To understand the nanosecond laser derusting process and reveal the derusting mechanism, the surface morphology and roughness of AH32 marine steel after laser derusting under different process parameters was studied. First, a nanosecond-pulse laser was used to clean the rust layer of the sample surface under different process parameters. Then, the surface roughness after cleaning was measured using laser scanning confocal microscopy, the micromorphology was measured by scanning electron microscopy, and an elemental analysis was performed using an energy dispersive spectrometer. Finally, the nanosecond laser derusting mechanism of AH32 marine steel was revealed combined with the experimental results. The experimental results indicated that the cleaned morphology was improved, and that the surface roughness was reduced under stepwise laser cleaning with energy densities of 30.6 and 10.2 J/cm2 at a scanning speed of 3 000 mm/s. Moreover, under the above cleaning conditions, the micromorphology of the matrix exhibited a micromelting state after laser derusting, in which the spot was smooth and uniform and the edge of the spot was distributed with a dendritic mastoid structure. It can be concluded that the use of a stepwise laser cleaning process can yield a better cleaning effect and higher cleaning efficiency. The derusting mechanism mainly includes hole blasting and melting-vaporization.

Phase-shifting locking is one of the key aspects of Scanning Beam Interference Lithography (SBIL) in achieving highly accurate exposure stitching of a large grating. To achieve this objective, a phase-shifting locking system was investigated for the stepping/scanning exposure trajectory of SBIL. Firstly, based on a previously proposed homodyne frequency-shifting interference pattern locking system and a heterodyne Littrow grating interferometer, a novel phase-shifting locking system scheme was proposed for SBIL. An experimental setup was then designed for this proposed scheme. Based on the setup, experimental and factor analyses were conducted to facilitate precise phase-shifting positioning control and an accuracy of ±3.27 nm (3σ, Λ=251 nm) was achieved. In addition, an accuracy of ±4.17 nm (3σ, Λ=251 nm) was achieved for phase-shifting locking control using a notch and PID control.

A high-bandwidth, two-degree-of-freedom nanopositioning stage based on the optimization of a flexible beam is proposed with the aim of improving the low-bandwidth performance, relative low-travel range, and poor coupling performance of the scanning positioning stage of Atomic Force Microscopy. Design optimization, simulation verification, and experimental analysis of the proposed stage are conducted as part of this process. Firstly, a parallel compliant moving stage composed of a doubly clamped beam and parallel hybrid beam is presented, while Castiglianos second theorem and Lagranges equation are applied to establish the mathematical model of its stiffness and natural frequency. Then, the maximum natural frequency and optimal size of the stage are obtained using optimization theory, while the optimization result reliability is verified using finite element method software. Finally, an experimental system is built and experiments are conducted on the developed stage. The experimental results indicate that the travel range of the proposed stage is 12.950 μm×13.517 μm, with a coupling error of less than 1.77%. The natural frequencies in the X and Y directions are 12.21 kHz and 13.50 kHz, respectively. In open loop, triangular waves with frequencies less than 1 kHz can be tracked well, effectively addressing the problems of slow response, small stroke, and poor coupling performance of the traditional scanning nanopositioning stage.

To meet the requirements of high dynamic characteristics and overcome the low quality of the main support structure, an optimized design was devised and test verification of the main support structure was conducted. First, to solve the problem that classic support structures could not be applied to this system, a general structural form combining a frame-type structure and a thin-walled tubular structure was proposed. This form conveniently integrates the hood and structure to improve quality and ensure rigidity. Second, multi-variable integration optimization was adopted to improve the first-order mode from 127 to 156 Hz when the quality requirement was also met, thereby effectively improving the dynamic characteristics. Thereafter, to determine the influence of the vibration environment on the system MTF, the influence of the vibration environment on the system MTF was analyzed by combining finite element analysis and a sensitivity matrix. The scope of application demonstrates that the structure could be applied to an airborne system with a pixel size of >10 μm. Finally, the effectiveness and feasibility of the design method and analysis process were verified by a vibration test and a wavefront aberration detection test. The optimized design method proposed in this paper can provide reference for the optimized design of airborne remote sensing instrument structures. This paper will contribute to the development of structure design techniques for airborne remote sensing instruments.

In this study, the hysteresis nonlinear law of piezoelectric ceramics was explored to provide a reference and theoretical basis for further correction of hysteresis nonlinearity. A hysteresis model was obtained through experimental measurements, and experimental data were analyzed. The corrected straight line of the rising trajectory was fitted; the voltage difference between the rising and falling trajectories at the same displacement was obtained as the compensation voltage; and the input voltage at the sampling points of the two trajectories was determined. The compensation voltage was found to satisfy an approximately polynomial relationship, where 5-, 10-, and 15-V asynchronous lengths were selected for driving experiments. Experimental results show that the rising trajectory (falling trajectory) polynomial parameters are R-square=0.999 2 (0.999 9), root mean square error (RMSE)=0.083 (0.080 86); R-square=0.999 7 (0.999 9), RMSE=0.057 39 (0.094 99); and R-square=0.995 2 (0.999 8), RMSE=0.291 6 (0.165 5). The experimental results also show that the R-square of the two trajectories is approximately 1; the RMSE is closer to 0; and the repeatability error is between 1.13% and 2.63%. The input and compensation voltages satisfied the approximately polynomial relationship, and the degree of fit and matching are high and repeated. The hysteresis model is proven to have good performance and predictability, thus providing a reference and theoretical basis to further correct the hysteresis nonlinearity of piezoelectric ceramic actuators.

To meet the new requirements for mobile robots in the field of modern engineering and expand the working opportunities of mobile robots, a spherical parallel leg mechanism of a wheel-leg composite mobile robot was proposed in this study. First, based on the closed-loop constraint equation and rotation transformation matrix of the spherical parallel leg mechanism, a mathematical model of its inverse position solution was constructed. Next, an analytical solution of the forward position solution of the spherical parallel leg mechanism was deduced by the algebraic elimination method. Then, the influence coefficient matrix of the velocity and acceleration of the spherical parallel leg mechanism was derived from the influence coefficient method. On this basis, the Lagrangian method was used. Kinematics and dynamics models were validated by numerical simulation. The maximum error between the given and calculated position data was 0.012 7 rad, and the error did not exceed 2.43% of the actual value. It is also found that the theoretical curve of the driving force of the spherical parallel leg mechanism coincided with the simulation curve of the virtual prototype. The error between the two was stable within a reasonable range of 0-1 N. The correctness of the kinematics and dynamics model is proved. The results from this study provide a theoretical reference for gait planning and motion control of a wheel-leg hybrid mobile robot.

To address the problem of mechanical resonance in the control system of an aeronautical optoelectronic stabilization platform, a novel structural filter based on a double-T notch filter was proposed in this study. The new structure filter, consisting of two double-T notches and a low-pass filter, could suppress both resonant and anti-resonant peaks while overcoming the problem whereby a single notch filter could not deal with anti-resonance. Compared to that of the double quadratic filter, the parameter adjustment of the structural filter was more flexible, and its peak size and width can be simultaneously adjusted. In addition, the experimental results show that the resonance curve fitted by the structural filter matches the actual sweep curve better. The stability accuracy of the control system is improved by 5.74 times and the closed-loop bandwidth is increased by 4.54 times after the structural filter is added to compensate the model. The new structure filter clearly aids in suppressing mechanical resonance and improving the performance of the control system.

Lunar calibration is an effective method for visible and near-infrared spectrum remote sensors. Original lunar observation data should be pre-processed into lunar full disk irradiance in order to compare them with a lunar radiation model. This paper describes the pre-processing of lunar data of the MERSI on the FY-3D satellite. The data pre-processing includes two steps. First, the lunar observation data are identified from the massive space view data. Then, the identified lunar original digital count is converted into full disk irradiance. Based on the MERSI observation mode, the lunar zenith and azimuth in the instrument reference coordinate system are used to construct a threshold model. When the zenith and azimuth meet the threshold conditions, the data are identified as lunar image. According to MERSIs imaging geometry and calibration formula, the lunar full disk irradiance can be calculated from two different lunar images. Irradiance from a single-detector multi-scan image needs to be corrected for the over-sample factor, and irradiance from a multi-detector single-scan image needs to be corrected for the radiation response difference between detectors. Based on this method, about 1 minute of lunar observation data among 30 days can be found and identified. The results show that the average difference of irradiance between the two methods is about 0.9%. The pre-processing method presented in this paper can find the original lunar observation data and convert them to full disk irradiance value, which provides a basis for further absolute radiation calibration and error analysis. It can also provide a reference for lunar calibration of other similar remote sensors.

A precise infrared target system (IRTS) plays a major role in an infrared imaging system. In this study, using the Zhang Zhengyou calibration method, we design a simple precision target based on the properties of a materials thermal and surface radiation. The design elements of the IRITS are presented according to the principles of the Zhang Zhengyou calibration method. Based on the mechanism of radiation, thermodynamic theory, and position relation of the feature position, a target plan possessing high and low radiation areas with sharp boundaries is designed. An error analysis of the target plan is then implemented according to the aging principle. Experiments are conducted with the designed precise IRTS. Results show that the designed IRTS can supply feature points by sharp boundaries from different radiation areas for an infrared camera and that the root mean square of the back-projection error is less than 0.16 pixels. These results suggest that the IRTS can be used to calibrate infrared cameras precisely.

Image style transfer exploits a specified style to modify given image content. An automatic image style transfer based on a Generative Adversarial Network (GAN) can reduce the workload and yield rich results. In some cases, the pair datasets required by the classical GAN were difficult to obtain. To overcome the limitations of paired datasets by a traditional GAN and improve the efficiency of style transfer, this study proposed an image style transfer method based on an improved Cycle-consistent adversarial network (CycleGAN). In this study, the deep residual network adopted by the conventional network generator was replaced by the dense connection convolution network, and a novel loss function composed of the same mapping and perceptual losses was used to measure the style transfer loss. These improvements were shown to increase the network performance, overcome the networks limitations on paired samples, and improve the quality of images generated by style migration. In addition, the stability was further improved and the network convergence speed was accelerated. Experiments demonstrate that the peak signal-to-noise ratio of the image generated by the proposed method increase 6.27% on average, where as the structural similarity index measure increased by approximately 10%. The improved CycleGAN image style transfer method proposed in this study can thus generate better style images.

To address the confusion and low accuracy of salient objects in co-saliency detection for image groups with complex environments, we proposed a co-saliency detection model based on objectness and a multi-layer linear model. First, we calculated the inter-saliency values using the background guidance factor weighted by saliency prior and objectness probability. We then designed a local region feature to calculate the intra-saliency values. The zero, first, and second Hu moments of the image were used to integrate the two-stage saliency values. Finally, saliency subgraphs were adaptively fused using a multi-layer linear model. Experimental results reveal that the AP scores of the proposed algorithm are 87.80% on iCoseg datasets and 83.50% on the MSRC dataset. Results from the evaluation of other experimental indicators are also improved significantly. The detected salient objects are more accurate and the adaptability of the algorithm is enhanced.

For the model reconstruction and pose estimation of non-cooperative rotating space targets with unknown model, the technology of graph-based optimization SLAM was applied to reduce the cumulative error in the pose tracking process by using 3D point clouds acquired through LiDAR. First, the relative pose between adjacent key frames was calculated by the Iterative Closest Point (ICP) algorithm, and the pose of the current key frame was obtained by the pose tracking method, thereby constructing the pose graph of the chaser spacecraft. Meanwhile, the global 3D signature GLAROT-3D (Geometric LAndmark relations ROTation-invariant 3D) was used to detect the loop closure, and adding the closed-loop constraint to the pose graph. Finally, the method based on pose graph optimization was used to adjust the pose and update the model point cloud. Experimental results show that in the simulation test, when the noise amplitude reaches 100mm, the attitude measurement error is less than 2°. In the field experiment, the attitude measurement error is less than 2.5°, and the target point cloud model is well reconstructed. Hence, the accuracy and the anti-noise ability of the proposed method can satisfy the mission requirements for the relative pose measurement of non-cooperative target.

In this study, to detect and evaluate welding defects effectively, magnetic flux leakage characteristics of magneto-optic imaging under alternating magnetic field excitation are proposed for application in the contour reconstruction of welding defects. A magnetic flux leakage reconstruction model was also established to study the two-dimensional contour characteristics of welding defects. First, based on the formation mechanism of a leakage magnetic field under an alternating magnetic field, the relationship between the two leakage magnetic field component signals (By and Bz) and the defect contour was discussed. Second, a generalized regression neural network was trained using numerical simulation data, to determine the model and to show that the leakage magnetic field signal can achieve defect contour reconstruction. Finally, the data derived from magneto optic imaging magnetic leakage characteristics were applied to the training of the model, to determine the feasibility of reconstruction. Experimental results show that the image data of the magnetic flux leakage characteristics are consistent with the contour reconstruction rule obtained through simulation, and a two-dimensional contour reconstruction of welding defects can be realized. Within a specific range, the greater (no less than 0.45 mm) the depth of the defect, the better the reconstruction effect.

With the improving detection sensitivity of star sensors, the number of stars in star catalogs has increased dramatically, reducing the speed and rate of star identification. To improve the identification speed and rate, a fast star identification algorithm based on multi-feature matching and built upon the foundation of the triangle algorithm was proposed. First, the preprocessed star catalog was partitioned using the inserted icosahedron of the celestial sphere. Then, the navigation feature library was constructed with the sides and the product of the radii of circumcircles and incircles as feature values. In addition, the navigation feature library was stored in blocks according to the hash function of the latter feature value. In the process of star identification, the product of the radii of the circumcircle and incircle in the observation triangle was used to rapidly locate the block of the navigation feature library, and then the observation triangle was identified in the block by using multi-feature matching. Finally, the sub-regions of the celestial sphere in the field of view were obtained, and then other navigation stars were identified in the sub-regions. The experimental results indicate that the identification performance of the proposed algorithm is related to the number of blocks. Based on a reasonable number of blocks, the proposed algorithm has advantages in identification speed and rate as well as in its robustness to star magnitude noise and false stars, compared with common triangle algorithms. The average identification time and rate of the proposed algorithm are 17.161 ms and 98.58%, respectively, which can meet the star sensor requirements for high identification speed and rate.

It is desirable to avoid the error and noise sensitivity of second-order tensor data leading to the disadvantages of low positioning accuracy with the Euler inversion method of magnetic object single-point positioning. For single-point positioning with only first-order tensor data, a method based on tensor derivative invariant relations is proposed. For the analysis of the magnetic dipole source tensor invariant and eigenvalue, two tensor derivative invariant relations were derived. The angle between the magnetic moment and position vector is constant and related to the tensor eigenvalue. The eigenvector of the absolute-minimum eigenvalue is perpendicular to the magnetic moment and position vector, and the eigenvectors of the remaining eigenvalues are coplanar with them. Thus, four possible solutions with respect to the quadrants of a plane above the magnetic source center were obtained, and the unique solution can be determined by the actual orientation and measured data. The results show that after error correction of the magnetic gradient tensor system, the positioning accuracy of a small-scale magnet (diameter of 5 cm, thickness of 0.5 cm) can be controlled within a root mean square error of 5 cm. Compared with the Euler inversion method, the proposed method has a greater detection distance with the same noise and exhibits a more reliable result.

To accurately determine the inner diameter contour of a microscopic image of a harness terminal for subsequent data analysis, a microscopic image segmentation algorithm based on variable exponential filtering is proposed in a chromaticity-brightness space. First, based on the color distribution characteristics of the microscopic image of a color wire harness terminal, color images in the red, green, and blue color space are converted to the chroma brightness space. In addition, lighting effects and separate luminance and chrominance information are eliminated, and chrominance information is converted to chroma spheres. Next, based on the total variation model, a variable exponential variational model on a chromaticity sphere is constructed to filter the chrominance information, where the variable exponential function is shown to have structural adaptive properties. Then, following analysis of the edge detection result of the monochrome channel, the Canny algorithm is used to perform edge detection in the red channel of the image. After the false discontinuous boundary curve is removed, the final inner diameter contour is obtained. Experimental results show that the circumference of the inner diameter of the terminal obtained by this method is compared with the manual measurement, and the deviation is less than 0.5%. The data regarding the inner diameter contour and circumference of the terminal obtained by the algorithm are also shown to be accurate.

In this study, to address the problem of limited adaptability and accuracy of the exact measuring angle method, which only calibrates the principal point and distortion in a one-dimensional direction with a line-scan camera, a two-dimensional high-precision calibration method for a line-scan camera was proposed. First, the inner orientation elements model of a line-scan camera was analyzed. Then, a two-dimensional calibration method based on a two-axis precise rotation stage was proposed for the model. Detailed calibration steps and data processing methods were also given. Finally, the calibration results of the proposed method were compared with those of the exact measuring angle method. The results show that the reprojection error of the calibration method proposed in this study is 0.34 pixels, which is a significant improvement in calibration precision when compared to the 1.25 pixels of the exact measuring angle method. The proposed calibration method does not require alignment, and the calibration process is simple.