Search Results (105)

Optical reflective markers are widely used in precision medicine, computer-assisted surgery, and robotic interventions. Nevertheless, intraoperative tracking still faces challenges such as sensor saturation, Point Spread Function (PSF) blooming, and flat-top artifacts, which affect localization precision and stability. Traditional deep learning detectors perform well in general object recognition but are limited in handling saturated infrared reflective markers due to their neglect of optical physics and inability to separate signal from blooming interference. This paper presents a physics-prior-guided network integrating a Brightness-Prior-Enhanced Spatial Attention (BPESA) mechanism for high-precision sub-pixel marker localization under saturation conditions. The method achieves a Root Mean Square (RMS) error of 0.52 pixels (approximately 0.11 mm) on a dataset of 8000 binocular images and reduces the localization error by approximately 54.4% compared with the baseline YOLOv8 model, while maintaining an inference speed of 134.6 FPS. The results demonstrate that optical blooming interference can be effectively mitigated by a learnable physics-prior branch, providing accurate marker coordinates that form a foundation for potential downstream tracking or navigation tasks. Full article

Background: Senior–Loken syndrome (SLS) is a rare autosomal recessive ciliopathy classically defined by the concurrence of nephronophthisis, frequently progressing to end-stage renal disease (ESRD), and retinal dystrophy, most commonly presenting as retinitis pigmentosa (RP). Given its phenotypic overlap with other renal–retinal syndromes, establishing a definitive diagnosis necessitates integrated clinical evaluation and molecular confirmation. Case Presentation: A 28-year-old Chinese female presented with a two-month history of binocular floaters. Her medical history was significant for ESRD of five years’ duration, managed with maintenance hemodialysis. Ophthalmic assessment revealed retinal pigment mottling along the inferior temporal arcades and generalized arterial attenuation. Spectral-domain optical coherence tomography demonstrated outer retinal thinning with loss of the ellipsoid zone at corresponding locations. Perimetry confirmed visual field constriction, and full-field electroretinography showed severely reduced rod- and cone-mediated responses. Genetic testing was performed and a pathogenic variant in the NPHP1 gene was identified. Segregation studies confirmed both parents as heterozygous carriers, consistent with autosomal recessive inheritance. Collectively, these findings established a diagnosis of SLS. Conclusions: This case reinforces that SLS should be considered in the differential diagnosis of any young patient exhibiting RP alongside chronic kidney disease, particularly in the setting of early-onset ESRD. It also illustrates the essential role of a coordinated, multidisciplinary approach—encompassing nephrology, ophthalmology, and genetics—in diagnosing complex ciliopathies. Genetic confirmation not only validates the clinical diagnosis but also provides a foundation for family counseling, prognostic stratification, and future eligibility for gene-specific therapeutic trials. Full article

Line-structured light three-dimensional (3D) measurement is commonly used for three-dimensional contour reconstruction of objects in complex industrial environments, but the problem of missing information occurs when three-dimensional reconstruction is performed on objects with smooth surfaces, single texture, and high reflectivity, resulting in defective reconstructed object surfaces. For this reason, this study proposes a fused complementary 3D reconstruction technique based on a polarization-based binocular line-structured light system. First, the reconstructed image of the object is captured using a Polarization Binocular Camera, and the polarized imaging effectively reduces the strong highlights and extracts more detailed information on the surface of the object. Then, the calibrated camera and optical planes are used to acquire the spatial coordinates of the object reconstructed by the left camera and right camera. Finally, the spatial coordinates obtained by the left camera and right camera are aligned, and the high-precision 3D reconstruction results are generated. The experimental results show that the proposed method can effectively improve the accuracy and robustness of 3D reconstruction, has a good application prospect, and can meet the technical requirements of industrial 3D measurement. Full article

Addressing the challenge of stable cloud velocity field estimation under complex sky conditions in ground-based cloud imaging, this paper proposes a comprehensive 3D cloud velocity calculation framework. The methodology integrates binocular stereo vision geometry, self-supervised deep feature learning, and graph attention-based matching. First, a self-supervised feature detection and description model tailored to the radiometric characteristics of cloud images is developed. By incorporating a homography adaptation strategy constrained by physical priors, the model acquires robust feature representations for weakly textured and highly deformable cloud masses without requiring labeled datasets. Subsequently, a Transformer-based graph neural network matcher is employed to establish global feature correspondences across both cross-view and cross-temporal dimensions, thereby substantially augmenting matching robustness. On this basis, the framework establishes a rigorous calibration model for fisheye cameras to derive cloud base height (CBH) via binocular geometry. These geometric constraints are then coupled with sequential feature tracking results to construct 3D velocity inversion equations, enabling an end-to-end mapping from 2D pixel coordinates to 3D physical space and providing direct estimation of physical cloud motion velocity in meters per second (m/s). The experimental results show that the proposed method extracts 4.5 times more feature points than the traditional SIFT method. Furthermore, the Pearson correlation coefficient for cloud motion trends in continuous sequences reaches 0.662 relative to baseline models, indicating good relative consistency in motion estimation. The framework achieves high-precision and stable velocity estimation across diverse cloud types, including cirrus, cumulus, stratus, and mixed clouds. Full article

To address the issue of low depth estimation accuracy in complex goji berry orchards, this paper proposes a method for identifying and locating goji berries that combines the YOLO-VitBiS object detection network with stereo vision technology. Based on the YOLO11n backbone network, the C3K2 module in the backbone is first improved using the AdditiveBlock module to enhance its detail-capturing capability in complex environments. The AdditiveBlock introduces lightweight long-range interactions via residual additive operations, thereby strengthening global context modeling without significantly increasing computation. Subsequently, a weighted bidirectional feature pyramid network is introduced into the Neck to enable more flexible and efficient feature fusion. Finally, a lightweight shared detail-enhanced detection head is proposed to further reduce the network’s computational complexity and parameter count. The enhanced model is integrated with binocular stereo vision technology, employing the CREStereo depth estimation algorithm for disparity calculation during binocular stereo matching to derive the three-dimensional spatial coordinates of the goji berry target. This approach enables efficient and precise positioning. Experimental results demonstrate that the YOLO-VitBiS model achieves a detection accuracy of $96.6\%$, with a model size of $4.3\mathrm{MB}$ and only $1.856\mathrm{M}$ parameters. Compared to the traditional SGBM method and other deep learning approaches such as UniMatch, the CREStereo algorithm generates superior depth maps under complex conditions. Within a distance range of 400 mm to 1000 mm, the average relative error between the estimated and actual depth measurements is $2.42\%$, meeting the detection and ranging accuracy requirements for field operations and providing reliable recognition and localization support for subsequent goji berry harvesting robots. Full article

Accurately measuring the verticality of a single pile is of crucial importance for ensuring the safe operation of offshore wind power projects. However, mainstream methods have disadvantages such as high dependence on manual labor, low real-time performance, and susceptibility to construction site conditions and marine environmental impacts. The method of measuring the verticality of a single offshore pile based on binocular vision is one of the emerging measurement methods, but there is currently a lack of research on measurement models. In order to clarify the principle of the method for measuring the verticality of a single pile at sea based on binocular vision, this paper starts from the imaging principle of the camera and studies and derives the measurement model of the verticality of a single pile in the global coordinate system and the error model of the measurement system. To verify the correctness of the model and method, a testing experimental platform was built to simulate the measurement of the ship under static and dynamic conditions, and the measurement results were compared with those of the total station. The experimental results show that in the static simulation experiment, the maximum absolute error of the verticality of a single pile is 0.2°, the maximum absolute error of the roll angle is 0.3°, and the maximum absolute error of the pitch angle is 0.3°. In the dynamic simulation experiment, the maximum absolute error of the verticality of a single pile is 0.4°, the maximum absolute error of the roll angle is 0.3°, and the maximum absolute error of the pitch angle is 0.3°. This paper verified the correctness of the model and provided model support for measuring the verticality of single piles at sea. Full article

The landing gear is one of the key components of an aircraft, enduring significant forces during takeoff and landing, and is influenced by various uncertain factors related to its structure. Therefore, conducting strength tests on the landing gear structure to study its ultimate load capacity is of great significance for structural design and analysis. This paper proposes a visual measurement method for dynamic pose of landing gear that combines stereo vision and CAD digital model. The method first establishes a measurement reference in CAD digital model and then uses close-range photogrammetry and binocular stereo vision technology to unify the coordinate system of the physical landing gear model with the measurement coordinate system of CAD model. Finally, during the motion of the landing gear, CAD model and the physical model can be synchronized by tracking a small number of key points, thus obtaining the complete motion state of the landing gear during the test. The experimental results demonstrate that the RMSE of the angle error is less than 0.1°, and the RMSE of the trajectory error is under 0.3 mm. This level of accuracy meets the requirements for pose measurement during the landing gear retraction and extension test. Compared to existing methods, this approach offers greater environmental adaptability, effectively reducing the impact of unfavorable factors such as occlusion during testing. It allows for the retrieval of pose information for any point on the landing gear, including its centroid. Full article

The size and mass of fish are crucial parameters in aquaculture management. However, existing research primarily focuses on conducting fish size and mass estimation under ideal conditions, which limits its application in actual aquaculture scenarios with complex water quality and fluctuating lighting. A non-contact size and mass measurement framework is proposed for complex underwater environments, which integrates the improved FishKP-YOLOv11 module based on YOLOv11, stereo vision technology, and a Random Forest model. This framework fuses the detected 2D key points with binocular stereo technology to reconstruct the 3D key point coordinates. Fish size is computed based on these 3D key points, and a Random Forest model establishes a mapping relationship between size and mass. For validating the performance of the framework, a self-constructed grass carp dataset for key point detection is established. The experimental results indicate that the mean average precision (mAP) of FishKP-YOLOv11 surpasses that of diverse versions of YOLOv5–YOLOv12. The mean absolute errors (MAEs) for length and width estimations are 0.35 cm and 0.10 cm, respectively. The MAE for mass estimations is 2.7 g. Therefore, the proposed framework is well suited for application in actual breeding environments. Full article

Accurate human keypoint localization in complex environments demands robust sensing and advanced modeling. In this article, we construct a ST-ConvLSTM network for 3D human keypoint estimation via millimeter-wave radar point clouds. The ST-ConvLSTM network processes multi-channel radar image inputs, generated from multi-frame fused point clouds through parallel pathways. These pathways are engineered to extract rich spatiotemporal features from the sequential radar data. The extracted features are then fused and fed into fully connected layers for direct regression of 3D human keypoint coordinates. In order to achieve better network performance, a mmWave radar 3D human keypoint dataset (MRHKD) is built with a hybrid human motion annotation system (HMAS), in which a binocular camera is used to measure the human keypoint coordinates and a 60 GHz 4T4R radar is used to generate radar point clouds. Experimental results demonstrate that the proposed ST-ConvLSTM, leveraging its unique ability to model temporal dependencies and spatial patterns in radar imagery, achieves MAEs of 0.1075 m, 0.0633 m, and 0.1180 m in the horizontal, vertical, and depth directions. This significant improvement underscores the model’s enhanced posture recognition accuracy and keypoint localization capability in challenging conditions. Full article

In response to the global labor shortage in the pear industry, the use of robots for harvesting has become an inevitable trend. Developing pear harvesting robots for orchard operations is of significant importance. This paper systematically reviews the progress of three key technologies in pear harvesting robotics: Firstly, in the field of recognition technology, traditional methods are limited by sensitivity to lighting conditions and occlusion errors. In contrast, deep learning models, such as the optimized YOLO series and two-stage architectures, significantly enhance robustness in complex scenes and improve handling of overlapping fruits. Secondly, positioning technology has advanced from 2D pixel coordinate acquisition to 3D spatial reconstruction, with the integration of posture estimation (binocular vision + IMU) addressing occlusion issues. Finally, the end effector is categorized based on harvesting mechanisms: gripping–twisting, shearing, and adsorption (vacuum negative pressure). However, challenges such as fruit skin damage and positioning bottlenecks remain. The current technologies still face three major challenges: low harvesting efficiency, high fruit damage rates, and high equipment costs. In the future, breakthroughs are expected through the integration of agricultural machinery and agronomy (standardized planting), multi-arm collaborative operation, lightweight algorithms, and 5G cloud computing. Full article

The precise locations of unknown non-cooperative targets are a long-standing technical problem that needs to be solved urgently in disaster relief and emergency rescue. An imaging model of photography to a non-cooperative target was established based on the binocular vision forward intersection. The collinear equation representing the spatial position relationship between the target and its two images was obtained through coordinate system transformation, and the system of equations to calculate the geographic coordinates of the target was derived, which realized the geo-location of the unknown non-cooperative target with no control points and no source. The composition and source of the error of this target location method were analyzed, and the equation to calculate the total error of the target location was obtained according to the error synthesis theory. The accuracy of the target location was predicted. When the elevation difference between the camera and the target is 3 km, the location accuracy is 15.5 m. The same ground target was imaged by a certain type of aerial camera at different locations 3097 m above ground, and a target location verification experiment was completed. The longitude and latitude of the target obtained were compared with the true geographic longitude and latitude, and the location error of the verification experiment was calculated to be 16.3 m. The research work of this paper provides a theoretical basis and methods for the precise locations of unknown non-cooperative targets and proposes specific measures to improve the accuracy of target location. Full article

To enable intelligent inspection of underground cable systems, this study presents a gecko-inspired quadruped robot that integrates multi-degree-of-freedom motion with a deep learning-based visual detection system. Inspired by the gecko’s flexible spine and leg structure, the robot exhibits strong adaptability to confined and uneven tunnel environments. The motion system is modeled using the standard Denavit–Hartenberg (D–H) method, with both forward and inverse kinematics derived analytically. A zero-impact foot trajectory is employed to achieve stable gait planning. For defect detection, the robot incorporates a binocular vision module and an enhanced YOLOv8 framework. The key improvements include a lightweight feature fusion structure (SlimNeck), a multidimensional coordinate attention (MCA) mechanism, and a refined MPDIoU loss function, which collectively improve the detection accuracy of subtle defects such as insulation aging, micro-cracks, and surface contamination. A variety of data augmentation techniques—such as brightness adjustment, Gaussian noise, and occlusion simulation—are applied to enhance robustness under complex lighting and environmental conditions. The experimental results validate the effectiveness of the proposed system in both kinematic control and vision-based defect recognition. This work demonstrates the potential of integrating bio-inspired mechanical design with intelligent visual perception to support practical, efficient cable inspection in confined underground environments. Full article

The 3D point cloud generated by MVSNet has good scene integrity but lacks sensitivity to details, causing holes and non-dense areas in flat and weak-texture regions. To address this problem and enhance the point cloud information of weak-texture areas, the BCA-MVSNet network is proposed in this paper. The network integrates the Bidirectional Feature Pyramid Network (BIFPN) into the feature processing of the MVSNet backbone network to accurately extract the features of weak-texture regions. In the feature map fusion stage, the Coordinate Attention (CA) mechanism is introduced into 3DU-Net to obtain the position information on the channel dimension related to the direction, improve the detail feature extraction, optimize the depth map and improve the depth accuracy. The experimental results show that BCA-MVSNet not only improves the accuracy of detail texture reconstruction, but also effectively controls the computational overhead. In the DTU dataset, the Overall and Comp metrics of BCA-MVSNet are reduced by 10.2% and 2.6%, respectively; in the Tanksand Temples dataset, the Mean metrics of the eight scenarios are improved by 6.51%. Three scenes are shot by binocular camera, and the reconstruction quality is excellent in the weak-texture area by combining the camera parameters and the BCA-MVSNet model. Full article

Aiming at the problems of visual occlusion, reduced positioning accuracy and pose loss in the dynamic scanning process of aviation large components, this paper proposes a binocular vision dynamic positioning method based on magnetic target. This method detects the spatial coordinates of the magnetic target in real time through the binocular camera, extracts the target center to construct a unified reference system of the measurement platform, and uses MATLAB simulation to analyze the influence of different target layouts on the scanning stability and positioning accuracy. On this basis, a dual-objective optimization model with the objectives of ‘minimizing the number of targets’ and ‘spatial distribution uniformity’ is established, and Monte Carlo simulation is used to evaluate the robustness under Gaussian noise and random frame loss interference. The experimental results on the C-Track optical tracking platform show that the optimized magnetic target layout reduces the rotation error of the dynamic scanning from 0.055° to 0.035°, the translation error from 0.31 mm to 0.162 mm, and the scanning efficiency is increased by 33%, which significantly improves the positioning accuracy and tracking stability of the system under complex working conditions. This method provides an effective solution for high-precision dynamic measurement of aviation large components. Full article