Weighted least-squares (WLS) phase unwrapping is widely used in optical engineering. However, this technique still has issues in coping with discontinuity as well as noise. In this paper, a new WLS phase unwrapping algorithm based on the least-squares orientation estimator (LSOE) is proposed to improve phase unwrapping robustness. Specifically, the proposed LSOE employs a quadratic error norm to constrain the distance between gradients and orientation vectors. The estimated orientation is then used to indicate the wrapped phase quality, which is in terms of a weight mask. The weight mask is calculated by post-processing, including a bilateral filter, STDS, and numerical relabeling. Simulation results show that the proposed method can work in a scenario in which the noise variance is 1.5. Comparisons with the four WLS phase unwrapping methods indicate that the proposed method provides the best accuracy in terms of segmentation mean error under the noisy patterns.

In phase-measurement profilometry (PMP), the gamma effect can cause severe nonlinear distortion of the phase pattern (i.e., water ripples on the surface profile). Gamma correction is an effective method to eliminate the gamma effects of commercial projectors. However, projector defocusing on the suppression of higher harmonics inevitably results in an estimated gamma deviation from the true value. In this study, gamma mapping is constructed using the duty ratio (DR) to code the chord distribution of the simulated distorted phase while considering projector defocusing. With the known gamma mapping, the accurate gamma is calculated by DR coding of the actual distorted phase under projector defocusing. Simulated experiments verified that the relative errors of the gamma calculated by the proposed method under different degrees of defocus were less than 3.5%. Furthermore, the experimental results demonstrate that the proposed gamma calculation method is robust to the defocus effect of the projector and that a smoother surface can be reconstructed after gamma correction.

This paper presents a gamma self‐correction method via chord distribution coding to eliminate the system's gamma effect. For this purpose, the prior gamma‐Gaussian characteristic value, which is extracted by chord distribution Gaussian fitting of the preset gamma's wrapped phase, is used to map the projector's gamma. Experimental results demonstrate that the proposed method has competitive accuracy and feasibility for fringe acquisition and gamma calculation. Thus, this method can be used for online gamma self‐correction in continuous three‐dimensional measurement.

When grating patterns are simultaneously projected by a dual-projection structured-light system, interference-like blur and brightness overexposure in the superposed area often cause miscalculation of the phase of the grating pattern. In this study, we proposed a novel method, to the best of our knowledge, that utilizes orthogonal grating encoding to retrieve the phases of superposed grating patterns. Specifically, we determined the frequency of the dual-projection pattern based on the condition that enabled the separation of superposed orthogonal signals in wireless communication. Additionally, the maximum intensity of the projected pattern was determined using the intensity-saturation relationship. By performing a discrete Fourier transform on a series of superposed grating patterns, we obtained the wrapped phase of the corresponding projected grating patterns in the space-time dimension. Finally, we reconstructed the measured object by fusing the point clouds obtained from the dual-projection structured-light system. The experimental results demonstrated that the encoded orthogonal grating patterns could eliminate interference-like blurring and brightness overexposure during superposition and obtain high-precision phase maps and 3D reconstruction results, which provides the possibility for the simultaneous reconstruction of multiprojection structured light.

Background: Oversize transport vehicles pose substantial challenges to vision-based traffic monitoring because their large bodies can obscure lane markings and wheel–ground contact regions, while elevated oblique camera views introduce strong perspective distortion. To address these difficulties, this study presents an occlusion-aware lane reconstruction and lane–vehicle mask-space reasoning framework for oversize-vehicle violation detection. The lane-perception branch, termed CLRNet-ALD, extends CLRNet through adaptive cross-layer feature fusion, large-kernel attention, deformable convolution, and multi-loss supervision to improve the reconstruction of partially or severely occluded lane boundaries. In parallel, a YOLOv12-based instance-segmentation branch extracts vehicle contours and instance masks. The lane and vehicle masks are subsequently aligned in a common image coordinate system, and interpretable geometric rules based on row-matched boundary comparison, instance counting, and mask overlap are used to identify designated-lane deviation, simultaneous passage of multiple oversize vehicles, and lane-line crossing. On a 4000-image lane-focused dataset, CLRNet-ALD increased precision, recall, F1-score, and IoU from 86.3%, 83.9%, 85.1%, and 71.4% to 90.2%, 88.5%, 89.3%, and 77.8%, respectively. In a separate 600-image system-level evaluation, 157 of 167 confirmed violation cases were correctly recovered, corresponding to a positive-case recovery rate of 94.0%, while 141 of 143 confirmed multi-type vehicle cases were correctly identified (98.6%). These results demonstrate the practical feasibility of combining occlusion-aware lane reconstruction with interpretable lane–vehicle spatial reasoning. Because the lane-component and system-level experiments were conducted at different development stages, the reported system-level results are interpreted as validation of the complete reasoning pipeline rather than as a controlled measurement of the end-to-end improvement attributable specifically to CLRNet-ALD.

The length of the irregular head and tail regions of medium and thick plates is a crucial factor affecting steel yield. This paper proposes a shearing length calculation method based on multi-modal edge feature fusion to address the limitations of current detection methods in shape recognition and length calculation accuracy. First, high-resolution industrial cameras capture the surface images of the steel plate, and motion distortion errors are corrected to accurately obtain the outer edge profile. Then, a dual-objective optimization model is designed, integrating width consistency criteria and edge linearity constraints, ensuring precise adaptive positioning of the shearing position. Finally, the shearing length of the irregular head and tail regions is calculated using the constructed cutting guide lines, and a polynomial regression error correction model is applied to improve calculation accuracy and stability, yielding the effective plate length. Experimental results show that the maximum error in detecting the shearing length of the head and tail is less than 1mm, with a 0.4% increase in steel utilization. The method has engineering application value by adjusting and optimizing rolling process parameters based on the shape and length of the irregular regions.