Stereoscopic Research Articles

A Comparative Study of the Modal Response of Additively and Subtractively Manufactured Thin Plates After Thermal Loading

Abstract Background Additively-manufactured parts contain residual stresses induced by manufacturing. These residual stresses can be relaxed or redistributed by thermal loading. The presence of internal stress influences the dynamic response of parts, and this is of particular interest in thin plates subject to thermoacoustic loading in hypersonic vehicles and fusion reactors. Objective To measure the changes in shape and modal frequencies caused by thermal loading of geometrically-reinforced thin plates that were additively manufactured in Inconel 625. Methods Plates were additively-manufactured in landscape and portrait orientations using laser powder bed fusion. The plates were heated to a nominal temperature of 820 ̊C, which was expected to alleviate the residual stress from the build process. Pre- and post-heating, their modal frequencies were found experimentally and pulsed-laser stereo (3D) digital image correlation was used to evaluate their modal shapes. The resultant modal frequencies and shapes were compared with those from a subtractively-manufactured plate. Results It was found that the heat cycle changed the shape of the plates relative to their as-manufactured state in addition to changing their natural frequencies and modal shapes. Conclusions The change in shape induced by heating caused shifts in the natural frequencies and changes in the corresponding modal shapes. The results show quantitatively for the first time the important role that residual stresses can play in the dynamic response of geometrically-reinforced thin plates manufactured by additive and subtractive processes.

Cross-scale backscattered-electron imaging and its application in revealing the microstructure-property relations

Scanning electron microscopy (SEM) has been widely utilized in the field of materials science due to its significant advantages, such as large depth of field, wide field of view, and excellent stereoscopic imaging. However, at high magnification, the limited imaging range in SEM cannot cover all the possible inhomogeneous microstructures. In this research, we propose a novel approach for generating high-resolution SEM images across multiple scales, enabling a single image to capture physical dimensions at the centimeter level while preserving submicron-level details. We adopted the SEM imaging on the AlCoCrFeNi2.1 eutectic high entropy alloy as an example. SEM videos and image stitching are combined to fulfill this goal, and the video-extracted low-definition images are clarified by a well-trained denoising model. Furthermore, we segment the macroscopic image of the eutectic high entropy alloy, and the area of various microstructures is distinguished. By combining the segmentation results and hardness experiments, we found that the hardness is positively correlated with the content of the body-centered cubic phase and negatively correlated with the lamella width. The whole procedure is also applied to a thermoelectric gradient material (PbSe-SrSe). Our work provides a feasible solution to generate macroscopic images based on SEM for further analysis of the correlations between the microstructures and spatial distribution, and can be widely applied to other types of microscopes.

Photometrically optimized event-based stereo 3D measurements.

Event cameras only report changes in brightness when thresholds in individual pixels relative to previous levels are crossed. They output sparse streams of events that quantify these changes spatially and temporally. We have developed a measurement system using two event cameras in a stereo configuration with a specialized projector for 3D measurements of static objects. We show that optimal contrast in the structured illumination yields high-fidelity measurements, reaching 3D root-mean-square deviations of 0.6 mm (≈0.3% of the measurement volume's diagonal). This work opens up the possibility of precise event-based, highly dynamic 4D measurements in the near future.

Exploiting commercial micro X-ray fluorescence systems for stereoscopic soft X-ray imaging

Background Commercial micro X-ray fluorescence (μXRF) systems often employ a tilted convergent beam, which can cause a misalignment between X-ray cartography and the corresponding visible images. This misalignment is typically considered a disadvantage, as it hinders the accurate spatial correlation of elemental information. However, this apparent drawback can be exploited to facilitate X-ray stereoscopy. Objective To demonstrate the use of unmodified commercial μXRF equipment to estimate the 3D configurations of metals and voids within a low-atomic-weight matrix, specifically polymethyl methacrylate, and to explore the implications for enhancing μXRF mapping techniques. This approach could have applications in materials science, archaeology, and other fields requiring non-destructive 3D chemical mapping. Methods Using unmodified commercial μXRF equipment, we leveraged both XRF and Compton scattering effects to quantitatively reconstruct the size, position, and depth of embedded tungsten, copper, and silver objects. The study specifically examines how beam divergence affects the acutance of objects located deeper within the sample. Results Our findings indicate a depth estimation bias ranging from 4% to 15% for depths between 24 mm, and a size estimation bias below 3.2%. These results validate the methodology and highlight the robustness of our approach under typical operational settings, suggesting that the technique could be applied to a wide range of samples with minimal modifications to existing μXRF systems. Conclusions The study confirms that the inclination-induced misalignment in μXRF systems can be harnessed to enhance three-dimensional imaging capabilities. Our work establishes a foundation for advancing current μXRF mapping techniques and interpretation strategies, potentially broadening the applications of μXRF in various scientific and industrial fields.

Exploration of Three-dimensional Modeling by Unmanned Aerial Vehicles for Application in Practical Engineering

Citing actual engineering cases, this paper aims to enhance the application of the model in actual engineering by describing the experience gained in the process of generating 3D stereo models generated by UAV aerial surveys, as well as the characteristics of the aerial photography and model generation process.

Automated Phenotypic Analysis of Mature Soybean Using Multi-View Stereo 3D Reconstruction and Point Cloud Segmentation

Phenotypic analysis of mature soybeans is a critical aspect of soybean breeding. However, manually obtaining phenotypic parameters not only is time-consuming and labor intensive but also lacks objectivity. Therefore, there is an urgent need for a rapid, accurate, and efficient method to collect the phenotypic parameters of soybeans. This study develops a novel pipeline for acquiring the phenotypic traits of mature soybeans based on three-dimensional (3D) point clouds. First, soybean point clouds are obtained using a multi-view stereo 3D reconstruction method, followed by preprocessing to construct a dataset. Second, a deep learning-based network, PVSegNet (Point Voxel Segmentation Network), is proposed specifically for segmenting soybean pods and stems. This network enhances feature extraction capabilities through the integration of point cloud and voxel convolution, as well as an orientation-encoding (OE) module. Finally, phenotypic parameters such as stem diameter, pod length, and pod width are extracted and validated against manual measurements. Experimental results demonstrate that the average Intersection over Union (IoU) for semantic segmentation is 92.10%, with a precision of 96.38%, recall of 95.41%, and F1-score of 95.87%. For instance segmentation, the network achieves an average precision (AP@50) of 83.47% and an average recall (AR@50) of 87.07%. These results indicate the feasibility of the network for the instance segmentation of pods and stems. In the extraction of plant parameters, the predicted values of pod width, pod length, and stem diameter obtained through the phenotypic extraction method exhibit coefficients of determination (R2) of 0.9489, 0.9182, and 0.9209, respectively, with manual measurements. This demonstrates that our method can significantly improve efficiency and accuracy, contributing to the application of automated 3D point cloud analysis technology in soybean breeding.

Impact of Phase Offset on the Flowfields Downstream of Synchronized Propellers

This paper reports on an experimental investigation of the interactions of two side-by-side propellers in hover (J=0) using stereoscopic particle image velocimetry (SPIV). The propellers employed were DJI 9443 propellers rotating at 4860 RPM with a spacing of 5% of the propeller diameter. The SPIV measurement images were phase-locked with the synchronized propeller rotation. Three experimental scenarios were considered: a) a single propeller; b) dual, counter-rotating propellers rotating exactly in phase; and c) dual, counter-rotating propellers rotating 90 deg out of phase. The propeller tip vortices, full-field velocity measurements, and overall momentum flux measurements were analyzed for the three scenarios. The interactions of adjacent propellers caused a 55% faster decay in the peak vorticity of the tip vortices and significantly altered the tip vortex trajectories when compared to a single-propeller scenario. There was negligible change in the magnitude of the peak vorticity in tip vortex cores, the dissipation of the tip vortices, and axial velocity flowfields downstream of the propellers for the two differing phase-offset dual-propeller configurations. For both dual-propeller scenarios, the average momentum flow rate was 2.2% higher than for the single-propeller case, while the peak value was 5.2% higher.

Blind Stereoscopic Omnidirectional Image Quality Assessment via a Binocular Viewport Hypergraph Convolutional Network

Blind Stereoscopic Omnidirectional Image Quality Assessment via a Binocular Viewport Hypergraph Convolutional Network

SpatialTouch: Exploring Spatial Data Visualizations in Cross-reality.

We propose and study a novel cross-reality environment that seamlessly integrates a monoscopic 2D surface (an interactive screen with touch and pen input) with a stereoscopic 3D space (an augmented reality HMD) to jointly host spatial data visualizations. This innovative approach combines the best of two conventional methods of displaying and manipulating spatial 3D data, enabling users to fluidly explore diverse visual forms using tailored interaction techniques. Providing such effective 3D data exploration techniques is pivotal for conveying its intricate spatial structures-often at multiple spatial or semantic scales-across various application domains and requiring diverse visual representations for effective visualization. To understand user reactions to our new environment, we began with an elicitation user study, in which we captured their responses and interactions. We observed that users adapted their interaction approaches based on perceived visual representations, with natural transitions in spatial awareness and actions while navigating across the physical surface. Our findings then informed the development of a design space for spatial data exploration in cross-reality. We thus developed cross-reality environments tailored to three distinct domains: for 3D molecular structure data, for 3D point cloud data, and for 3D anatomical data. In particular, we designed interaction techniques that account for the inherent features of interactions in both spaces, facilitating various forms of interaction, including mid-air gestures, touch interactions, pen interactions, and combinations thereof, to enhance the users' sense of presence and engagement. We assessed the usability of our environment with biologists, focusing on its use for domain research. In addition, we evaluated our interaction transition designs with virtual and mixed-reality experts to gather further insights. As a result, we provide our design suggestions for the cross-reality environment, emphasizing the interaction with diverse visual representations and seamless interaction transitions between 2D and 3D spaces.

Optic nerve assessment with stereo photographs and ultra-widefield scanning laser ophthalmoscope images

ABSTRACT Clinical relevance Clinical evaluation of the optic nerve using 3-D stereo disc photographs is considered the gold standard for estimating vertical cup-to-disc ratios. Ultra-widefield retinal imaging has gained increasing popularity to document and screen the health of the retina and optic nerve. Background Glaucoma is often first identified or suspected based on initial optic nerve assessment. Despite technological advancements in imaging, stereo disc photography remains the gold standard of optic nerve assessment. This study compared vertical cup-to-disc ratio (VCDR) estimations using ultra-widefield scanning laser ophthalmoscopy images with estimations using paired stereo disc images. Methods Thirty-five healthy and 35 glaucomatous eyes were imaged using Zeiss VISUCAM® PRO NM and Optos Monaco® (optomap®) [Registered in U.S. Patent and Trademark Office] devices. Four experienced clinicians assessed each image in a masked fashion and graded the VCDR. Stereo viewers and review software tools were used to determine VCDRs. Bland-Altman plots and regression/correlation analyses were used to check VCDR agreement. Results Bland-Altman analyses showed low mean differences between devices with VCDR estimation for each grader (−0.03, +0.01, −0.02, and +0.03). Difference vs. means plots suggest stable differences across the range of measurement. Conclusions VCDR estimations were comparable using optomap® images and VISUCAM® PRO NM stereo optic disc images.

A novel 3D image registration technique for augmented reality vision in minimally invasive thoracoscopic pulmonary segmentectomy.

In this feasibility study, we aimed to create a dedicated pulmonary augmented reality (AR) workflow to enable a semi-automated intraoperative overlay of the pulmonary anatomy during video-assisted thoracoscopic surgery (VATS) or robot-assisted thoracoscopic surgery (RATS). Initially, the stereoscopic cameras were calibrated to obtain the intrinsic camera parameters. Intraoperatively, stereoscopic images were recorded and a 3D point cloud was generated from these images. By manually selecting the bifurcation key points, the 3D segmentation (from the diagnostic CT scan) was registered onto the intraoperative 3D point cloud. Image reprojection errors were 0.34 and 0.22 pixels for the VATS and RATS cameras, respectively. We created disparity maps and point clouds for all eight patients. Time for creation of the 3D AR overlay was 5min. Validation of the point clouds was performed, resulting in a median absolute error of 0.20mm [IQR 0.10-0.54]. We were able to visualize the AR overlay and identify the arterial bifurcations adequately for five patients. In addition to creating AR overlays of the visible or invisible structures intraoperatively, we successfully visualized branch labels and altered the transparency of the overlays. An algorithm was developed transforming the operative field into a 3D point cloud surface. This allowed for an accurate registration and visualization of preoperative 3D models. Using this system, surgeons can navigate through the patient's anatomy intraoperatively, especially during crucial moments, by visualizing otherwise invisible structures. This proposed registration method lays the groundwork for automated intraoperative AR navigation during minimally invasive pulmonary resections.

Mitigation of AI adoption bias through an improved autonomous AI system for diabetic retinal disease

Where adopted, Autonomous artificial Intelligence (AI) for Diabetic Retinal Disease (DRD) resolves longstanding racial, ethnic, and socioeconomic disparities, but AI adoption bias persists. This preregistered trial determined sensitivity and specificity of a previously FDA authorized AI, improved to compensate for lower contrast and smaller imaged area of a widely adopted, lower cost, handheld fundus camera (RetinaVue700, Baxter Healthcare, Deerfield, IL) to identify DRD in participants with diabetes without known DRD, in primary care. In 626 participants (1252 eyes) 50.8% male, 45.7% Hispanic, 17.3% Black, DRD prevalence was 29.0%, all prespecified non-inferiority endpoints were met and no racial, ethnic or sex bias was identified, against a Wisconsin Reading Center level I prognostic standard using widefield stereoscopic photography and macular Optical Coherence Tomography. Results suggest this improved autonomous AI system can mitigate AI adoption bias, while preserving safety and efficacy, potentially contributing to rapid scaling of health access equity. ClinicalTrials.gov NCT05808699 (3/29/2023).

Assessing Stereo Camera Applicability for Digital Monitoring of Cattle Exterior

The paper highlights the substantial potential for digitalization in animal husbandry. Current applications of digital technologies include replacing manual data collection on animal phenotypes, particularly linear measurements of physical traits. In creating a contactless digital monitoring system for cattle exterior traits, cameras play a crucial role, as they enable accurate distance measurement to the object. Digital reconstruction of animal body morphometry using a contactless measurement method and automated size determination can effectively address the issues with inaccuracy and subjectivity associated with traditional scoring methods. (Research purpose) The study aims to explore the feasibility of using stereo cameras to measure object distances with the required accuracy and to analyze the performance of the stereo vision system across different areas of the frame. (Materials and methods) The study used a stereo pair of two 1/3-Inch CMOS OV4689 4-megapixel lenses mounted on the board, spaced at 6.3 centimeters from each other. Accurate distance measurement was considered achieved when the error remained within 1-2 percent (1-2 centimeters) of the object's distance (0.5-1 meter). A marked sheet with 25 centimeter intervals served as a test stand, and the stereo cameras captured the stand from distances of 30 to 100 centimeters, with a 10 centimeter increments. (Results and discussion) The study employed two camera configurations over two stages: a single stereo camera and a block of three cameras. Filming results with the single stereo camera showed a measurement error of 5-10 centimeters at distances ranging from 0.3 to 1 meter from the object. For the three-camera block, the accuracy remained comparable. It was found that accuracy was higher at the center of the frame, with an average error of 3 centimeters at viewing angles near zero.(Conclusions) The study confirmed that the number of stereo pairs does not impact accuracy, and the observed error represents the accuracy limit for these stereo pairs in stereo vision applications.

Monitoring Indian ungauged small reservoirs volume from remote sensing: Feasibility, bias and perspectives

What remote sensing products can be used to better quantify the water stored in hundreds of thousands Indian Small Reservoirs (SR)? This ungauged hydrological component of the water cycle is intermittently filled with rainwater runoff, constantly reshaped by farmers since last two decades, crucial for upstream irrigated agriculture. Given the small size and shallow depth of those reservoirs, usual remote sensing techniques (Altimeters and LIDAR) used in spatial hydrology to monitor their water level are not adapted. We evaluated the uncertainty of SR volume retrieval methods based on surface water estimates from Sentinel-2 and associated volumes from global available DEM at a medium to coarse resolution. Four pair of stereoscopic images at Very High Resolution (VHR) from Pléiades satellites were acquired during the last two dry hydrological years (2016 and 2019), when SR were totally empty. The Pléiades DEMs produced were cross validated with LIDAR IceSAT-2 products, and used to extract 504 SR bathymetries within an area covering 1,813 km2 located in the Telangana state (114,789 km2). We compared Pléiades based retrievals to freely available regional to global DEM to explore the regional volume retrieval Bias: ALOS World 3D-30 m, WorldDEM GLO-30 at 30 m TanDEM-X DEM at 90 m and one Indian DEM (CartoDEM at 30 m). The Mean Absolute Percentage Error (MAPE) of reservoir volumes from global DEMs range from 47% to 78%. MAPE are 17%, 29% and 46% for Pléiades DEM resampled at 12, 30 and 90 m, respectively. In a near future, upcoming stereoscopy satellite missions at lower costs and with larger coverage and shorter revisit such as CO3D will provide 12m or higher resolution DEMs that, if acquired in dry years, will lead to acceptable MAPE (< 20%), to monitor empty SR geometries throughout India and other semi-arid areas in the world.

Trait-focused low-cost and stereo-based 3D plant modeling for phenotyping via deep neural detection

Abstract. Phenotyping, the measurement of plants physical traits, plays a pivotal role in advancing sustainable agricultural practices. Therefore, developing efficient, low-cost, means to generate such measures is vital. Though image based 2D-driven methods are commonly applied for that purpose due to their processing simplicity, it is clear that only 3D information can offer the necessary plant details. Notwithstanding, the generation of such data is challenged by the requirement to acquire a large set of images or the use of active sensors, which exhibit sensitivity to illumination and require lengthy acquisition campaigns. Consequently, 3D plant phenotyping is presently limited to controlled laboratory conditions and is hardly applied in actual growth setups. To address this shortcoming, this paper argues that by focusing on relevant plant traits, modeling can be simplified and the need for detailed plant reconstruction can be relieved. Accordingly, only a minimal set of images, specifically a stereo pair, can suffice for the reconstruction, thereby providing a low-cost sensing solution. To facilitate the reconstruction, we adapt an anchor-free detection deep neural network and integrate low- and high-level features to accurately detect our plant traits of interest. As the paper demonstrates our adapted network facilitates a robust 3D reconstruction of the entities of interest. Performance analysis demonstrates how our detection is reliable and accurate compared to standard anchor-free frameworks, translating to accurate reconstruction, as we validate against 3D plant scans.

FA-MSVNet: multi-scale and multi-view feature aggregation methods for stereo 3D reconstruction

FA-MSVNet: multi-scale and multi-view feature aggregation methods for stereo 3D reconstruction

Volcanic ring morphostructures in Momchilgrad area of Southeastern Rhodope, Bulgaria

The focus of this study is the interpretation of images obtained by the ASTER optical instrument on the TERRA satellite. Its data is processed and presented in stereoscopic images. Anaglyphic glasses with red and cyan color filters are used to obtain the realistic 3D effect. The structural analysis of data shows two ring morphostructures with distinct patterns of radial and concentric lineaments forming a clear circular morphology with a diameter of up to 20 km, located in Eastern Rhodope, eastern from the town of Momchilgrad, Bulgaria. Several volcanic edifices of intermediate composition and many acid extrusions have been mapped. These relatively young structures are directly reflected in the modern morphostrucural plan and can be easily detected in aerial photographs and space images.

Visual Field Progression in the Ocular Hypertension Treatment Study.

To determine the rate of visual field (VF) loss before and after the diagnosis of primary open angle glaucoma (POAG) in the Ocular Hypertension Treatment Study (OHTS). Pre-specified analyses of data collected prospectively in a clinical trial with extended follow-up. Participants who developed POAG during OHTS 1 and 2 (February 1994 to December 2008) constitute an inception cohort. Follow-up data were collected in OHTS 3 (January 2016 to April 2019). Analyses were performed from July 2021 to August 2022. In OHTS 1 and 2, visual field (VF) tests were performed every 6 months and stereoscopic optic disc photographs were taken every 12 months. These tests were repeated in OHTS 3. Slopes of mean deviation (MD) were calculated by linear regression for all eyes in OHTS 1 and 2: eyes that did not develop POAG, eyes that developed optic disc POAG only, and eyes that developed VF POAG with/without optic disc POAG. Mean pre- and post-POAG slopes were calculated for eyes with a minimum of 5 VFs for each period. Mean age at diagnosis of POAG was 66.4 ± 9.5 SD years (n= 282 participants), (56%) were male, 61% were White non-Hispanic and 32% were Black not Hispanic by self-report. The post-POAG slope was -0.40 ± 0.64 SD dB/year for all POAG eyes (n=280 eyes), -0.19 ± 0.4 SD dB/yr. for optic disc POAG only eyes (n=112 eyes), and -0.54 ± 0.7 SD dB/yr. for VF POAG eyes with or without optic disc POAG (n=168 eyes). Among the VF POAG eyes, 69 (41%) had post-POAG MD slopes worse than or equal to -0.5 dB/year, 35 (21%) had slopes worse than or equal to -1.0 dB/year, and 9 (5.4%) had slopes worse than or equal to -2.0 dB/year. Some participants in OHTS had rapid rates of VF loss in one or both eyes despite being followed in a clinical study. This emphasizes that ocular hypertensive patients require careful follow-up, especially those at high risk of developing POAG to ensure early diagnosis and appropriate treatment of POAG.

Experimental study of flow dynamics in a dual-stage annular swirling jet

In this study, three-dimensional flow structures at a Reynolds number of 16 000 are measured by stereoscopic particle image velocimetry to reveal the dynamic structures of a dual-stage co-rotating annular swirling jet with a blunt separating wall. The swirler number of the outer swirling jet is fixed at around 0.5 and that of the inner swirling jet is varied from 0 to 0.7. Spectral proper orthogonal decomposition is used for extracting organized structures and the topological evolution from a spatiotemporal flow field. The evolution of mean-flow topography is initially depicted as the flow transition from the center body wake to the presence of a central recirculation zone, with an increase in the inner swirl number varying from 0 to 0.7. Two precessing vortex cores (PVCs) with different frequencies are observed in the presence of a central recirculation zone, and the temporal independence of the two PVCs is identified. A transition region exists between the two PVCs because of the different dominant axial regions of the two PVCs, which is located between the central recirculation induced by inner swirling and that by the merged swirling flow. The conservation of circulation for both PVCs was confirmed in the transition region, and the two PVCs exhibited independent single-helical modes. Furthermore, the main frequencies of the two PVCs are proportional to the inner swirl number; however, they are higher than those of the corresponding single swirling jet. As predicted by the Landau equation, both PVCs had the same critical swirl number, suggesting that the two structures occurred simultaneously.

Large-scale and very-large-scale motions in a moderate-Reynolds-number turbulent boundary layer over a flat plate towed in a tank

An experimental work is conducted to investigate large-scale motions (LSMs) and very-large-scale motions (VLSMs) of a turbulent boundary layer generated by a flat plate towed in a tank. Reynolds number based on friction velocity and boundary layer thickness covers a range of 1200 ≤Reτ≤ 3100. The streamwise–wall-normal plane and the streamwise–spanwise plane are separately measured using time-resolved stereoscopic particle image velocimetry (TR-SPIV) and time-resolved planar PIV (TR-PIV), respectively. The large-scale coherent structures in the two orthogonal planes are identified by the clustering method. Based on spatiotemporal evolutions of instantaneous flow fields, it is observed that low- and high-speed coherent structures collide and lean against each other, generating a strong shear layer between them. The low-speed coherent structures move away from the wall, while the high-speed coherent structures travel toward the wall near the interaction regions during the shearing, which may lead to merging and splitting of the coherent structures. The coherent structures grow in streamwise scales by self-stretching and coalescence with adjacent coherent structures. The properties of the LSMs and VLSMs, including scales, inclination angles, meandering behaviors, and contributions of high- or low-speed VLSMs to skin friction, are quantitatively examined. Furthermore, the dynamic characteristics of the LSMs and VLSMs are classified into merging, splitting, growing, and shortening, with occurrence frequencies independent of Reynolds number. Effects of shearing on the merging and splitting events are discussed. Results from this work suggest that the primary formation mechanism of the VLSMs is the self-growing of LSMs vigorously stretching along the streamwise direction.