As a key part of the fabric quality control for the textile industry, it is important to detect fabric defects quickly, accurately, and efficiently. To address the missed detection of the defects with tiny, extreme aspect ratios, and low contrast in fabric images, an improved YOLOv8s-HCG algorithm is proposed in this paper. First, the histogram specification algorithm is used to enhance the defect feature expression of low-contrast fabric images at the input side. Second, the content-aware reassembly of features (CARAFE) operator is used instead of the nearest-neighbor interpolation operator for up-sampling in the YOLOv8s. The CARAFE operator aggregates contextual information in a large receptive field to reassemble abundant detailed features, and reassembles features by using targets in the feature map with the adaptive reassembled kernel. Finally, the global attention mechanism module is added into the YOLOv8s neck network to construct the interdependence relationship between fabric defect image channels and spatial dimensions to capture important features. The algorithm was validated on a self-made fabric dataset and two public fabric datasets. The comparison experimental results show that the detection performance of the proposed algorithm in this paper is better than the other algorithms for defects with extreme aspect ratios, tiny, and low contrast. This research is of great significance to the fabric defect detection industry.

Frequency-modulated continuous-wave (FMCW) lidar offers high precision and the capability for non-cooperative target sensing. However, its performance is critically limited by laser frequency modulation nonlinearity. Conventional resampling methods for nonlinearity correction necessitate an auxiliary interferometer path delay greater than four times the target distance, as dictated by the Nyquist theorem. This requirement results in bulky and instability-prone systems. The work proposed a novel resampling method based on the Chinese remainder theorem (CRT) to resolve the fundamental trade-off between precision and range. The proposed approach digitally subdivided the phase of a short auxiliary interferometer signal into multiple co-prime factors, constructing several equivalent undersampled sequences. The true distance was mapped into a set of unique, shorter residue distances, which were then uniquely reconstructed via the CRT. The method achieved an over 55-fold extension of the unambiguous range (from 0.91 to 50.16 m) using a short auxiliary path of 3.669 m. A measurement precision was maintained better than 15.0 µm in 8 m. A superlinear range extension was achieved without proportional hardware or data overhead, providing a robust and compact solution for high-precision, long-range FMCW metrology.

To address the challenge of achieving high-precision registration between visible light(RGB)and photoluminescence(PL)images in Micro LED defect inspection,which arises from substantial modality dif-ferences,this study introduces a robust multimodal image registration approach capable of attaining sub-pixel accuracy,aiming to establish a direct mapping between the physical structure and electrical characteristics of the chips.We propose a registration method that integrates structural feature constraints with bidirectional re-sidual optimization.First,leveraging the geometric regularity of Micro LED arrays,a tailored feature detec-tion strategy is employed:electrode centers in RGB images are accurately extracted via ellipse fitting and Density-Based Spatial Clustering of Applications with Noise(DBSCAN),while chip centers in PL images are localized using an enhanced watershed algorithm with sub-pixel refinement.Second,during the registration optimization stage,a bidirectional residual constraint framework is constructed,incorporating a confidence weighting mechanism derived from residual distribution analysis.The optimal affine trans-formation parameters are then estimated using an iterative reweighted least squares method.Experimental results demonstrate that the proposed method achieves sub-pixel-level accuracy,with a mean absolute error(MAE)of 0.823 pixels,representing a 94.2%reduction compared to baseline methods.The root mean square error(RMSE)is 0.996 pixels,the maximum error remains below 2.839 pixels,and the inlier rate at-tains 75.0%.Each registration process takes only 0.036 seconds on average,achieving an order-of-mag-nitude improvement in computational efficiency over traditional mutual information(MI)methods.By effect-ively mitigating feature mismatch and outlier interference in multimodal images,the proposed method out-performs conventional approaches in terms of registration accuracy,robustness,and efficiency,thereby providing a reliable technical foundation for precise defect detection and multimodal analysis of Micro LED chips.

Deep silicon-etched microstructures are essential in both micro-electro-mechanical systems (MEMS) devices and advanced integrated-circuit packaging. Accurately acquiring the depth non-destructively is critically required. This work proposes a flexible multi-mode spectral interferometry applicable to microstructures of diverse geometry. First, the spectral reflectance (SR) mode is suited for narrow and high-aspect-ratio microstructures, requiring the light spot to be bigger than the width (width ≤10μm, aspect ratio ≥10:1). Second, the optical coherence tomography-like mode is more effective for microstructures with moderate or low aspect-ratios (width 10μm to 30μm, aspect ratio ≤10:1). Third, the image positioning-based spectroscopy (IP spectroscopy) mode is optimal while the light spot is smaller than the width. For demonstration, microstructures with widths ranging from 1 to 50μm, depths from 12 to 546μm, and aspect ratios between 1.2:1 and 51:1 are measured by the homemade apparatus. This method provides a metrology solution for multi-scale microstructure in MEMS and 3D-ICs.

Frequency-modulated continuous wave ranging technology is widely used in industrial manufacturing. Because the on-site working environments are complex, the measured target may have slight vibration. The Doppler frequency shift caused by vibration leads to the broadening of the beat frequency signal generated by the superposition of the reference light wave and the measuring light wave in the continuous wave ranging system, which reduces the measurement accuracy of the ranging system. In this paper, we analyze the measurement principle of frequency modulation continuous wave ranging technology and the influence of target vibration on the range measurement accuracy. The analysis results indicate that the target vibration displacement can amplify the measurement error by dozens to hundreds of times. To address the above measurement error caused by tiny vibration displacement, in this paper we propose a vibration suppression method for the frequency modulation continuous wave ranging based on the four-wave mixing effect. Firstly, the generation principle of four-wave mixing is introduced. The single-frequency laser is used as the pump light, and the tunable laser is used as the signal light. These two lights are simultaneously incident into the highly nonlinear fiber. The converted light is generated by the third-order parametric process of the nonlinear medium of the fiber. The converted light and the signal light from the tunable laser form a symmetrical light source with scanning directions that are completely opposite. When the superimposed upper and lower scanning lights are filtered by high-pass filtering, the influence of vibration on the measurement signal is suppressed. Secondly, an experimental system of the four-wave mixing frequency modulation continuous wave is built, and the single-point measurement stability is verified. Based on the Mach-Zehnder interferometric measurement principle, a four-wave mixing effect frequency modulation continuous wave range measurement system is constructed. The static target at 6.9 m away is measured by this constructed ranging system. The distance of peak-to-bottom range is reduced from 199.8 mu m before vibration suppression to 16 mu m, which represents an improvement of more than 12 times. A ranging accuracy comparison experiment is also carried out in a range of 6-7.2 m, and the ranging accuracy is lower than 9.4 mu m. Experimental results demonstrate that the vibration suppression method utilizing the four-wave mixing effect can effectively improve the measurement accuracy of frequency modulation continuous wave ranging, which holds significant importance in industrial applications.

Frequency-sweep amplitude-modulation laser ranging (FSAMLR) is a ranging method that determines the target distance by solving for the in-phase frequency, characterized by high measurement accuracy and low system complexity. To address issues such as the low signal-to-noise ratio in sampled waveforms containing in-phase frequencies and the resulting limitations in solving accuracy, a method based on singular spectrum analysis combined with local parabolic fitting (SSA-LPF) is proposed. The principle of FSAMLR is outlined, emphasizing that ranging accuracy depends on the precision of the in-phase frequencies. Subsequently, simulations compare the solving accuracy of in-phase frequencies among the swing method, parabolic fitting, cubic fitting, and quartic fitting, using identical sampled waveforms filtered via the SSA method. Parabolic fitting is verified to enhance solution accuracy. Simulation results demonstrate that parabolic fitting achieves a 95.7% reduction in mean absolute deviation relative to the swing method and a 65.6% improvement over other least-squares fitting methods. Experimental analysis indicates that the SSA-LPF method yields a ranging standard deviation below 30 mu m across varying distances and sweep steps. Adopting the SSA-LPF method in FSAMLR enhances ranging efficiency while maintaining high ranging accuracy.

Abstract A systematic, fast, and precise tracking method with strong adaptive capability is proposed. Based on a theoretical analysis of the control strategy and mathematical model of the permanent magnet synchronous motor, a feed-forward loop is actively introduced into the servo control system, which can reduce the steady-state tracking error to nearly zero even when the target moves at high speed and/or with a high constant and/or time-variant acceleration rate. Considering the moving randomness and uncertainty of the handheld laser measurement target, the characteristic and deficiency of the traditional “current “ statistical (CS) model are analyzed from theory and practice, then an improved ‘current’ statistical (ICS) model with stronger adaptability is proposed to estimate the target motion state in real-time. The estimated angle velocity and angle acceleration of the maneuvering target are calculated by the Kalman filtering method using the range (distance from the center of the laser tracking gimbal to the target), rotated angle, miss distance, and other parameters, and the estimated information is fed forward to the closed speed control loop. A compound tracking control structure of three feedback closed loops (electric current, speed, and position loops) and an equivalent feed-forward loop was constructed. The compound method was validated through a series of tracking experiments. Based on the ICS model and compound control structure, the one-dimensional steady-state tracking error of the target is almost always zero when the target is moving at an acceleration of 10 m s −2 or in the sinuous motion (−0.4cos(5 t)) state. The two-dimensional steady-state tracking errors were both within 120 μm when the target made a uniform circular motion with an acceleration rate of 25 m s −2 or 36 m s −2 . These results demonstrate that the adaptive and precise tracking performance is greatly improved, and the proposed method is better than state-of-the-art methods. This indicates that this method can better meet the high tracking accuracy requirements of the laser tracking system and can provide a useful reference for large-scale equipment docking, dynamic tracking, and laser communication.

In dynamic and complex environments, absolute distance measurement offers advantages for the manufacturing and maintenance of high-end precision equipment. This paper proposes a frequency-sweep amplitude-modulation laser ranging method. This method inherits the merits of frequency-sweep polarization-modulation ranging method in industrial applications. It avoids requiring a single modulator to perform double modulation, and obviates the necessity to account for electromagnetic interference-induced perturbations to the polarization state of the ranging system. Furthermore, the proposed ranging system boasts a compact architecture and exhibits exceptional integration potential. The least squares method is employed to acquire the integral amplitude waveform, and the target distance is determined by identifying the minimum values and its corresponding in-phase frequencies. The validity of the method is verified through theoretical derivation and simulation. An experimental system is designed and constructed, with ranging parameters optimized via simulation and experimental test. At target distances of approximately 20 m, 35 m, and 50 m, with sweep step sizes of 10 kHz and 100 kHz, the ranging stability is confirmed as 9-30 mu m and the accuracy as +/-(25-60)mu m.

The precise acquisition of distance data critically depends on Absolute Distance Measurement (ADM) technology. This paper comprehensively examines frequency-modulated continuous wave (FMCW), pulsed laser ranging, and ultrasonic ranging technologies, evaluating their application value across automotive manufacturing, aerospace, and industrial automation. FMCW technology demonstrates superior ranging accuracy and resolution through frequency difference analysis despite its susceptibility to multipath interference; pulsed laser ranging achieves long-distance precision via optical pulse transmission while facing ambient light limitations; ultrasonic technology offers cost-effective short-range solutions though constrained by environmental factors. Current challenges include signal processing complexity in FMCW systems, environmental sensitivity of pulsed laser ranging measurements, and resolution limitations in ultrasonic methods. This paper summarizes the research results of the past few years, points out the problems existing in the current technology, and puts forward suggestions for future research directions, providing an important reference for research in related fields.

To address the conflict between precision and cost in existing circular grating self-compensation methods—where high accuracy requires numerous reading heads while reducing their quantity compromises angular measurement precision—this paper proposes a harmonic self-compensation method for relative errors based on a hybrid reading-head layout. The approach employs a Hybrid Layout of Two-uniformly-distributed (DUA) and Three-uniformly-distributed (TUA) reading heads (HTD Layout). By synergistically leveraging the harmonic error suppression characteristics of both layouts across different orders, it achieves full-bandwidth error self-compensation except for 6th-order and multiple harmonics. The core procedure involves: first, deriving relative errors by calculating differences between the DUA average and common reading head, thereby filtering out true values while preserving specific-order error components; second, extracting 3rd-order harmonic coefficients from these relative errors for compensation; finally, implementing real-time compensation by superimposing 3rd-order harmonics onto the TUA average to correct specific 3rd-order error harmonics. Experimental results demonstrate significantly enhanced angular measurement precision, with peak-to-valley error reduced from 160″ to 2.2″ while achieving substantial accuracy improvements with fewer hardware components

To address the limitations of six-degree-of-freedom (6-DOF) visual tracking systems in terms of low dynamic performance and the inability to perform measurements under light occlusion, this paper proposes a novel laser tracking attitude dynamic measurement method integrating vision and inertial measurement unit (IMU). A tailored real-time calibration model is developed to enable precise alignment of the vision and IMU coordinate systems during dynamic operations. To overcome the limitations of extended Kalman filter (EKF) that rely on static noise assumptions in complex dynamic measurement scenarios, this study proposes an adaptive mechanism based on multi-factor influence analysis. Through dynamic monitoring of visual observation noise, an adaptive EKF algorithm is designed to enhance performance under varying conditions. When integrated into a vision/IMU fusion system, the algorithm significantly enhances the system's sensitivity to variations in visual sensor observation noise caused by changes in lighting and motion states. This capability allows rapid adjustment of filtering parameters, leading to substantial improvements in filtering accuracy and stability. Simulation results demonstrate that the proposed algorithm outperforms traditional EKF in fusion measurement accuracy when tested in simulated noise environments featuring sinusoidal and random jump signals. Furthermore, the fusion measurement accuracy is mainly influenced by the data density of the visual sensor and the measurement distance. Finally, the proposed method was validated on a precision turntable. Experimental results reveal that at a 3 m measurement distance and within a 0 degrees-25 degrees angular range, when the turntable rotates at a constant speed of 5 degrees/s along a single axis, the maximum absolute deviation in repeatability of the fused attitude measurements was below 0.11 degrees, and the system is capable of achieving a high measurement frequency of 100 Hz.

Using a single GHz-rate electro-optic comb in synthetic wavelength interferometry, we perform absolute distance measurement with a ±4-µm precision and 3.7-m non-ambiguity range. Three comb lines from the source are employed for this purpose.

Currently, the measurement of roll of laser tracking cooperative target using vision-based methods often necessitates the presence of multiple feature points on the target, distributed according to specific rules. However, this requirement undermines the inherent characteristic of laser tracking which is to consistently track a single point. In order to simplify the target structure by leveraging this characteristic, this paper investigates a visual measurement method for determining the roll of a cooperative target based on laser tracking. Firstly, this paper describes the characteristics of visual target detection for laser-tracked cooperative targets using active infrared detection and analyses the principles behind visual measurements based on distributed feature points in laser-tracked cooperative targets. Secondly, this paper constructs a measurement model for determining the roll of a cooperative target using active infrared detection and features, and establishes a corresponding measurement system. Finally, experiments are conducted and accuracy is analysed. The result demonstrates that this proposed method only requires a simple structure and achieves an error better than ±0.08° when only one active infrared feature point is present, thereby meeting medium- to high-precision application requirements.

We demonstrate high-precision absolute distance measurement through cascaded synthetic wavelength interferometry (SWI) based on a single GHz-rate electro-optic frequency comb (EO comb). For this purpose, 3 of its discrete comb modes are selectively filtered by cascaded dense wavelength division multiplexers to serve as phase-linked continuous-wave (CW) lasers with relative frequency stability inherited from the microwave frequency reference. We vary the frequencies of these comb modes to generate a series of synthetic wavelengths by periodically switching the EO comb repetition rate. This allows us to obtain a maximum non-ambiguity range (NAR) of ∼3.74 m accompanied by the instrument’s ± 4 µm precision validated by a commercial HeNe interferometer. In addition to static targets, a velocity measurement with < ±15 µm/s residual compared to the HeNe interferometer is also shown. Our simple and robust technique for high-precision absolute distance and velocity measurement is well-suited for many industry-relevant problems requiring the combination of micrometer precision with a meter-scale NAR.

The laser tracker plays an important role in large-scale, high-accuracy measurement fields due to its exceptional accuracy and strong dynamic performance. For dynamic measurement tasks, using a laser tracker network measurement system is both effective and efficient. In current research on the laser tracker network, once the reference coordinate system is unified, raw observations from each station are used directly without adjustment, which limits further improvements in dynamic measurement accuracy. To address this issue, a dynamic measurement adjustment method based on distance constraints is proposed. First, four fixed points are placed on the moving target, and the distances between these points are computed using the trilateration network adjustment. These distances are then used as constraints for adjusting dynamic measurement data over time. In the user manual for Leica AT960/AT930 laser trackers, the static measurement accuracy is specified according to the ASME B89.4.19-2006 standard. However, no explicit specification is provided for dynamic measurement accuracy. To address this, the paper proposes a method for evaluating the dynamic measurement accuracy of a networked laser tracker system. Finally, a dynamic measurement system was established using a synchronization trigger and four laser trackers, and experiments were conducted with a tetrahedral artifact as the target. Experimental results show that the dynamic measurement accuracy improves by more than 70% after adjustment.

Acousto-optic frequency shifter (AOFS) is the key component in optical metrology, encompassing applications such as length, velocity, and frequency measurements. The acoustooptic crystal suffers from thermal accumulation during operation, leading to significant changes in its optical and physical properties. As a result, the diffracted beam of an AOFS undergoes phase and frequency drift, degrading the measurement precision of optical metrology systems. In this letter, we evaluate the path length and optical frequency fluctuation of the 1(st)-order diffracted beam passing through an AOFS due to thermal effect. Our experiment reveals linear correlation between both path length drift rate and frequency drift versus temperature drift rate. The measured proportionality factor are 0.791 mu m/K and 1.019 Hz/K/s, respectively. Under thermal equilibrium conditions, AOFS-contributed noise floor in distance, velocity and frequency measurement are similar to 100 pm at 100 ms averaging time, 2.5 mu m/s at 1 ms averaging time, and 1.56 x 10(-20) at 10,000 s averaging time, respectively.

Laser-based light detection and ranging (LiDAR), which requires both high precision absolute distance measurement and velocity measurement, has been widely applied in various fields including surface profiling, remote sensing, imaging, etc. Optical frequency combs, which contain numerous frequency-stabilized comb modes, have been superior candidates as laser sources of LiDARs [1]. By stitching the comb modes, frequency modulated comb LiDAR is able to achieve simultaneous measurement of absolute distance and velocity [2]. However, the distance measurement precision of this method is usually at the magnitude of millimeter or sub-millimeter, which is insufficient for many applications. In this work, we demonstrate simultaneously absolute distance measurement and velocity measurement with high precision based on a 25.3 GHz electro-optic frequency comb (EO comb). We select three comb modes from a single EO comb to generate cascaded synthetic wavelengths that enable absolute distance measurement with <1 μm precision and 1.48 mm non-ambiguity range (NAR). The NAR could be easily extended by slightly changing the repetition rate. Simultaneous velocity measurement is realized by recording the Doppler frequency shift of the comb modes.

To solve the problem of measuring the azimuth attitude angle of a laser tracking cooperative target at a long distance, a high-precision measurement method based on beam direction is proposed in this paper. First, the optical properties of the cube-corner prism and the principle of laser tracking based on the cube-corner prism are described in detail. Second, this paper proposes a method for azimuth attitude angle measurement based on beam direction using the cube-corner prism with an incision, a scanning mirror, and a photodetector. It then analyzes the mathematical relationship between the laser beam vector, the cube-corner prism with incision, the scanning mirror, the photodetector, and the cooperative target, and establishes the attitude solution model with a cooperative target design. Finally, a high-precision attitude measurement system is constructed, experimental data acquisition and parameter model calibration are carried out, and accuracy analysis and evaluation are carried out through verification experiments. The experimental results show that the measuring device proposed in this paper is ingenious and practical, and the measuring angle error is less than 0.0187° in the working range of ±20°, which meets the demand of high-precision engineering applications.

The laser tracker network measurement system can expand the measurement range and improve precision while leveraging the benefits of a single laser tracker. Establishing a unified spatial datum is a core requirement for constructing a laser tracker network measurement system. Traditionally, rotation and translation parameters are determined using the triangulateration network adjustment model which is constrained by the precision of the laser tracker's angular observations, making further improvement in adjustment precision challenging. To address this issue, this paper proposes a method based on the unified least squares adjustment combining two adjustment models: the trilateration network,which uses only distance observations, and the triangulateration network,which uses both distance and angle observations. Firstly, the coordinates of the orientation points in the triangulateration network are treated as virtual observations, and the cofactor matrix of these points is calculated using the high-precision trilateration network adjustment model. Then angle and distance observations are combined with the virtual observations to form new observations. Finally the adjustment is conducted according to the least squares principle. Simulation and experimental results indicate that the proposed method improves the precision of rotation and translation parameters by over 30% compared to the traditional method.

The triangulateration network adjustment method is a fundamental algorithm for constructing a measurement system composed of multiple laser trackers. This method is based on a prior-weight-matrix model considering varied accuracies of angle and distance observations from laser trackers, which can successfully reduce the influence of angle errors on the adjustment results. However, the precision of the measurement system based on the triangulateration network adjustment method is considerably reduced when gross errors are present in the observations using the least squares principle. This paper proposes a robust estimation method using raw observation data with different prior precision from laser trackers. First, the standard deviation of unit weight for each type of raw observation data is computed separately using the median function. Then, the raw observation data are reweighted using the Institute of Geodesy and Geophysics III (IGG3) equivalent weight function. Simulation and experimental results show that the proposed method achieves higher precision compared to the traditional robust estimation method when the observations contain data with different units and prior precisions.