Depth Plane Research Articles

Chatter Detection in Milling of Carbon Fiber-Reinforced Composites by Improved Hilbert-Huang Transform and Recurrence Quantification Analysis.

In the paper, the problem of chatter vibration detection in the milling process of carbon fiber-reinforced plastic is investigated. Chatter analysis may be considered theoretically based on data from impact test of an end mill cutter. However, a stability region obtained in such way may not agree with the real one. Therefore, this paper presents a method that can predict chatter vibrations based on cutting force components measurements. At the beginning, a stability lobe diagram is created to establish the range of experimental test in the plane of tool rotational speed and depth of cut. Next, an experiment of composite milling is performed. The experimentally-measured time series of cutting forces are decomposed with the use of the improved Hilbert–Huang transform (HHT). To detect chatter, statistical methods and recurrence quantification analysis (RQA) are used. However, much better results are obtained when new chatter indexes are proposed. The indexes, derived directly from the HHT and RQA methods, can be used to build an effective chatter prediction system.

Foveated near-eye display using computational holography

Holographic display is the only technology that can offer true 3D with all the required depth cues. Holographic head-worn displays (HWD) can provide continuous depth planes with the correct stereoscopic disparity for a comfortable 3D experience. Existing HWD approaches have small field-of-view (FOV) and small exit pupil size, which are limited by the spatial light modulator (SLM). Conventional holographic HWDs are limited to about 20° × 11° FOV using a 4 K SLM panel and have fixed FOV. We present a new optical architecture that can overcome those limitations and substantially extend the FOV supported by the SLM. Our architecture, which does not contain any moving parts, automatically follows the gaze of the viewer’s pupil. Moreover, it mimics human vision by providing varying resolution across the FOV resulting in better utilization of the available space-bandwidth product of the SLM. We propose a system that can provide 28° × 28° instantaneous FOV within an extended FOV (the field of view that is covered by steering the instantaneous FOV in space) of 60° × 40° using a 4 K SLM, effectively providing a total enhancement of > 3 × in instantaneous FOV area, > 10 × in extended FOV area and the space-bandwidth product. We demonstrated 20° × 20° instantaneous FOV and 40° × 20° extended FOV in the experiments.

Experience with searching in displays containing depth improves search performance by training participants to search more exhaustively

In a typical visual search task, participants search for single targets amongst displays containing non-overlapping objects that are presented on a single depth plane. Recent work has begun to examine displays containing overlapping objects that are presented on different depth planes to one another. It has been found that searching displays containing depth improves response accuracy by making participants more likely to fixate targets and to identify targets after fixating them. Here we extended this previous research by seeking first of all to replicate the previous pattern of results, and then to determine whether extensive training using depth in search transfers to two-dimensional displays. We provided participants with sixteen sessions of training with displays containing transparent overlapping objects presented in depth, and found a similar pattern of results to our previous study. We also found evidence that some performance improvements from the depth training transferred to search of two-dimensional displays that did not contain depth. Further examinations revealed that participants learn to search more exhaustively (i.e., search for longer) in displays containing depth. We conclude that depth does influence search performance but the influences depend very much on the stimuli and the degree of overlap within them.

Depth Plane Separation Affects Both Lightness Contrast and Assimilation.

Lightness contrast and assimilation are two opposite phenomena: contrast occurs when a gray target perceptually acquires a complementary color than the bordering, inducing, surfaces; assimilation is when a gray target perceptually acquires the same color component as the inducers. Previous research has shown that both phenomena are affected by the manipulation of depth between the inducers and target. However, different results have been reported; it is not clear whether contrast persists when inducers are non-coplanar with the target. Previous studies differ for the spatial configuration of the stimuli and the technique adopted to manipulate depth. The aim of this research was to measure the effects of manipulating the depth between inducers and target in comparable conditions. Results show that contrast persists, but largely reduces, after depth manipulation while assimilation reverses to contrast. Furthermore, interesting asymmetries between white and black inducers emerged with white inducers favoring contrast and black inducers favoring assimilation. These results provide further evidence that high-level processes of visual processing are involved in both phenomena, with important consequences for lightness theories.

Crowding Effects Across Depth are Fixation-Centered for Defocused Flankers and Observer-Centered for Defocused Targets.

Depth needs to be considered to understand visual information processing in cluttered environments in the wild. Since differences in depth depend on current gaze position, eye movements were avoided by short presentations in a real depth setup. Thus, allowing only peripheral vision, crowding was tested. That is, the impairment of peripheral target recognition by the presence of nearby flankers was measured. Real depth was presented by a half-transparent mirror that aligned the displays of two orthogonally arranged, distance-adjustable screens. Fixation depth was at a distance of 190 cm, defocused depth planes were presented either near or far, in front of or behind the fixation depth, all within the depth of field. In Experiments 1 and 2, flankers were presented defocused, while the to-be-identified targets were on the fixation depth plane. In Experiments 3–5, targets were presented defocused, while the flankers were kept on the fixation depth plane. Results for defocused flankers indicate increased crowding effects with increased flanker distance from the target at focus (near to far). However, for defocused targets, crowding for targets in front of the focus as compared to behind was increased. Thus, defocused targets produce decreased crowding with increased target distance from the observer. To conclude, the effects of flankers in depth seem to be centered around fixation, while effects of target depth seem to be observer-centered.

Visual Search Fixation Strategies in a 3D Image Set: An Eye-Tracking Study

AbstractIn this study, we explore whether the inclusion of monocular depth within a pseudo-3D picture gallery negatively affects visual search strategy and performance. Experimental design facilitated control of (i) the number of visible depth planes and (ii) the presence of semantic sorting. Our results show that increasing the number of visual depth planes facilitates efficiency in search, which in turn results in a decreased response time to target selection and a reduction in participant average pupil dilation—used for measuring cognitive load. Furthermore, results identified that search strategy is based on sorting, which implies that an appropriate management of semantic associations can increase search efficiency by decreasing the number of potential targets.

68‐1: Investigation on Defocusing‐Induced Accommodation Shift in Microlens Array‐Based Near‐Eye Light Field Displays

The accommodation response in a microlens array‐based near‐eye light field display is found to considerably shift towards the central depth plane but not at the reconstructed depth plane due to defocusing of the lens array. The shifts are characterized and can be fully addressed by compensating in a binocular system.

A multimedia speech corpus for audio visual research in virtual reality (L).

Virtual reality environments offer new possibilities in perceptual research, such as presentation of physically impossible but ecologically valid stimuli in contrived scenarios. To facilitate perceptual research in such environments, this study presents a publicly available database of anechoic audio speech samples with matching stereoscopic and 360° video. These materials and accompanying software tool allow researchers to create simulations with up to five talkers positioned at arbitrary azimuthal locations, at multiple depth planes, in any 360° or stereoscopic environment. This study describes recording conditions and techniques, contents of the corpus, and how to use the materials within a virtual reality environment.

P‐204: Late‐News Poster: Stereoscopic AR Displays ‐ Towards Solid‐State Multi‐Focal Architecture

In this work an entirely solid‐state multi‐focal display architecture for augmented reality near‐eye displays is introduced and analyzed. At the core of the solution is volumetric display module with discrete image depth planes as means of mitigating vergence‐accommodation conflict in stereoscopic systems and enabling safe near‐3D‐content output. The physical image depth planes are realized as switchable liquid‐crystal based optical diffuser elements. Our analysis includes evaluation of the role of a cell‐gap and spacers on a resultant image quality.

A Novel Multi-Focus Image Fusion Network with U-Shape Structure.

Multi-focus image fusion has become a very practical image processing task. It uses multiple images focused on various depth planes to create an all-in-focus image. Although extensive studies have been produced, the performance of existing methods is still limited by the inaccurate detection of the focus regions for fusion. Therefore, in this paper, we proposed a novel U-shape network which can generate an accurate decision map for the multi-focus image fusion. The Siamese encoder of our U-shape network can preserve the low-level cues with rich spatial details and high-level semantic information from the source images separately. Moreover, we introduce the ResBlocks to expand the receptive field, which can enhance the ability of our network to distinguish between focus and defocus regions. Moreover, in the bridge stage between the encoder and decoder, the spatial pyramid pooling is adopted as a global perception fusion module to capture sufficient context information for the learning of the decision map. Finally, we use a hybrid loss that combines the binary cross-entropy loss and the structural similarity loss for supervision. Extensive experiments have demonstrated that the proposed method can achieve the state-of-the-art performance.

Self-Blinking Dyes Unlock High-Order and Multiplane Super-Resolution Optical Fluctuation Imaging.

Most diffraction-unlimited super-resolution imaging critically depends on the switching of fluorophores between at least two states, often induced using intense laser light and specialized buffers or UV radiation. Recently, so-called self-blinking dyes that switch spontaneously between an open, fluorescent "on" state and a closed, colorless "off" state were introduced. Here, we exploit the synergy between super-resolution optical fluctuation imaging (SOFI) and spontaneously switching fluorophores for 2D and 3D imaging. SOFI analyzes higher order statistics of fluctuations in the fluorophore emission instead of localizing individual molecules. It thereby tolerates a broad range of labeling densities, switching behavior, and probe brightness. Thus, even dyes that exhibit spontaneous blinking characteristics that are not suitable or suboptimal for single molecule localization microscopy can be used successfully for SOFI-based super-resolution imaging. We demonstrate 2D imaging of fixed cells with almost uniform resolution up to 50-60 nm in 6th order SOFI and characterize changing experimental conditions. Next, we investigate volumetric imaging using biplane and eight-plane data acquisition. We extend 3D cross-cumulant analysis to 4th order, achieving super-resolution in 3D with up to 29 depth planes. Finally, the low laser excitation intensities needed for single and biplane self-blinking SOFI are well suited for live-cell imaging. We show the perspective for time-resolved imaging by observing slow membrane movements in cells. Self-blinking SOFI thus provides a more robust alternative route for easy-to-use 2D and 3D high-resolution imaging.

Augmented reality display device with continuous depth rendering capabilities

True 3D display with continuous depth rendering is a key issue for augmented reality display system. However, the discrete sampling characteristics of current display devices makes the 3D images discontinuous in both lateral and depth directions. For depth information, very limited depth planes can be rendered, especially when the resolution of the display device is low. To address the limitation, a kind of augmented reality display device that can output continuous depth information using a modulated illumination integral imaging technique is demonstrated. Light field with a continuous directional component is achieved in a digitized integral imaging system by using a modulated convergent backlight. A custom-designed transparent off-axis spherical reflective lens is used as an optical combiner to project the light field images into the real world. Two prototype augmented reality display devices are fabricated that provided continuous depth information ranging from 0.5 m to 3 m for 3D images in front of the optical combiner. 3D images with continuous focus cues and double depth of field are realized using a simple optical structure. The imaging mechanisms of the augmented reality display devices are deduced analytically. The imaging characters of the augmented reality display devices are tested and verified.

Visual perception in a bottlenose dolphin (Tursiops truncatus): Successful recognition of 2-D objects rotated in the picture and depth planes.

Aquatic species such as bottlenose dolphins can move in 3 dimensions and frequently view objects from different orientations. This study examined their ability to identify 2-D objects visually despite changes in orientation across 2 rotation planes. A dolphin performed a matching-to-sample task in which a sample was presented at a different orientation from its match in a 3-alternative choice array. Samples were presented at 6 aspect angles in the picture plane (0°, ±45°, ±135°, 180°) and 6 aspect angles in the depth plane (0°, -45°, ±90°, +135°, 180°). Alternatives were always presented at 0°. Performance was significantly better than chance for all aspect angles in both rotation plane tests. There was a significant linear decline in accuracy as the sample object was rotated from 0° toward 180° in the picture plane. Performance with familiar objects (M = 97.1%) exceeded performance with novel objects (M = 76.9%). In the depth plane rotation test, there was a significant quadratic trend in accuracy as the sample object was rotated from 0° toward 180°, in which performance was significantly lower at ±90° than at all other orientations. Performance in the picture plane where all object features were available irrespective of orientation was significantly better than performance in the depth plane where the availability of visible features were dependent upon orientation (M = 81.2% vs. M = 63.0%). The dolphin's performance in this study shows evidence of both viewpoint-independent and viewpoint-dependent processes. (PsycInfo Database Record (c) 2020 APA, all rights reserved).

Enhanced 3D Point Cloud from a Light Field Image

The importance of three-dimensional (3D) point cloud technologies in the field of agriculture environmental research has increased in recent years. Obtaining dense and accurate 3D reconstructions of plants and urban areas provide useful information for remote sensing. In this paper, we propose a novel strategy for the enhancement of 3D point clouds from a single 4D light field (LF) image. Using a light field camera in this way creates an easy way for obtaining 3D point clouds from one snapshot and enabling diversity in monitoring and modelling applications for remote sensing. Considering an LF image and associated depth map as an input, we first apply histogram equalization and histogram stretching to enhance the separation between depth planes. We then apply multi-modal edge detection by using feature matching and fuzzy logic from the central sub-aperture LF image and the depth map. These two steps of depth map enhancement are significant parts of our novelty for this work. After combing the two previous steps and transforming the point–plane correspondence, we can obtain the 3D point cloud. We tested our method with synthetic and real world image databases. To verify the accuracy of our method, we compared our results with two different state-of-the-art algorithms. The results showed that our method can reliably mitigate noise and had the highest level of detail compared to other existing methods.

A 3D Image Quality Assessment Method Based on Vector Information and SVD of Quaternion Matrix under Cloud Computing Environment

With the increasing demands of end-users to the visual perception in three-dimension (3D) image, quality assessment for 3D imageis dominantly required as the feedback information for multimedia transmission systems. In this paper, a novel full-reference quality assessment method by considering the depth and integral color information of 3D image under cloud computing environment is proposed. Based on the property of the depth information in 3D image, the depth map is firstly separated into different planes according to the perception of human visual system (HVS). Then, after express the image pixels of every separated plane through quaternions, the structural and energy information are separated by quaternion singular value decomposition (QSVD). The distortion of structural and energy in every plane are calculated in various formulas respectively. The final result is calculated in terms of the global score, which synthesizes the structural and energy distortion scores in every individual depth plane. It should be pointed out that the chrominance information is employed in our mechanism to evaluate the color image quality because of its useful characteristic for 3D color image quality assessment, and its spatial correlation is used for calculating structural distortion through vector cross-product. Our experimental results confirm that the proposed method has achieves better performance under cloud computing environments compared with other existing 3D image quality assessment methods.

Depth plane adaptive integral imaging system using a vari-focal liquid lens array for realizing augmented reality.

Augmented reality (AR) is an interactive experience of a real-world environment where the objects that reside in the real world are enhanced by computer-generated perceptual information. Despite its attractive features, AR has not become popular because of the visual fatigue that many people face when they experience it. Many methods have been introduced to solve this visual fatigue problem and one of these methods is an integral imaging system that provides images almost continuous viewpoints and full parallax. However, the integral imaging system, which uses a lens array with a fixed focal length, has limited depth of focus (DOF) range. As a result, images that are outside of the DOF range become distorted. In this paper, a vari-focal liquid lens array was fabricated and the optical characteristics of the lens array were evaluated. Using the vari-focal liquid lens array, the DOF range was extended and high-resolution images are realized without restriction of depth range in an AR system.

A Depth-Dependent Integrated VF Simulation for Analysis and Visualization of Glaucomatous VF Defects.

PurposeVisual fields (VF) are measured monocularly at a single depth, yet real-life activities require people to interact with objects binocularly at multiple depths. To better characterize visual functioning in clinical vision conditions such as glaucoma, analyzing visual impairment in a depth-dependent fashion is required. We developed a depth-dependent integrated VF (DD-IVF) simulation and demonstrated its usefulness by evaluating DD-IVF defects associated with 12 glaucomatous archetypes of 24-2 VF.MethodsThe 12 archetypes included typical variants of superior and inferior nasal steps, arcuate and altitudinal defects, temporal wedge, biarcuate, and intact VFs. DD-IVF simulation maps the monocular 24-2 VF archetypes to binocular ones as a function of depth by incorporating three parameters of fixation, object, and interpupillary distances. At each location and depth plane, sensitivities are linearly interpolated from corresponding locations in monocular VF and returned as the higher value of the two.ResultsThe simulation produced 144 DD-IVFs for multiple depths from combinations of 12 glaucomatous archetypes. The DD-IVFs are included as a Shiny app in the binovisualfields package. The number of impaired locations in the DD-IVFs varied according to the overlap of VF loss between eyes.ConclusionsOur DD-IVF program revealed binocular functional visual defects associated with glaucomatous archetypes of the 24-2 pattern and is designed to do the same for empirically measured VFs. The comparison of identified visual impairments across depths may be informative for future empirical exploration of functional visual impairments in depth in glaucoma and other conditions leading to bilateral VF loss.Translational RelevanceOur DD-IVF program can reveal depth-dependent functional visual defects for clinical vision conditions where 24-2 test patterns are available.

Global depth perception alters local timing sensitivity.

Dynamic environments often contain features that change at slightly different times. Here we investigated how sensitivity to these slight timing differences depends on spatial relationships among stimuli. Stimuli comprised bilaterally presented plaid pairs that rotated, or radially expanded and contracted to simulate depth movement. Left and right hemifield stimuli initially moved in the same or opposite directions, then reversed directions at various asynchronies. College students judged whether the direction reversed first on the left or right–a temporal order judgment (TOJ). TOJ thresholds remained similar across conditions that required tracking only one depth plane, or bilaterally synchronized depth planes. However, when stimuli required simultaneously tracking multiple depth planes–counter-phased across hemifields–TOJ thresholds doubled or tripled. This effect depended on perceptual set. Increasing the certainty with which participants simultaneously tracked multiple depth planes reduced TOJ thresholds by 45 percent. Even complete certainty, though, failed to reduce multiple-depth-plane TOJ thresholds to levels obtained with single or bilaterally synchronized depth planes. Overall, the results demonstrate that global depth perception can alter local timing sensitivity. More broadly, the findings reflect a coarse-to-fine spatial influence on how we sense time.

Electrically high-resistance liquid crystal micro-lens arrays with high performances for integral imaging 3D display

In this paper, we propose a high-resistance (Hi-R) liquid crystal (LC) micro-lens arrays (MLAs) with a low operating voltage and a controllable focal length for integral imaging (II) 3D display. The focal lengths of the Hi-R LC MLAs can be effectively controlled by applying different operating voltages and driving frequencies. The experimental results indicate that the focal lengths can be adjusted from 4.27 mm to 2.88 mm when the operating voltage applied to the Hi-R LC MLAs increases from 1.8 Vrms to 2.8 Vrms at the driving frequency of 130 kHz. Compared with the II 3D display by using ITO circular-hole electrodes-based or Al circular-hole electrodes-based LC MLAs without Hi-R layer, the II 3D display based on the Hi-R LC MLAs with Al circular-hole electrodes can reveal clear 3D scenes at the corresponding center depth plane (CDP) when a low operating voltage is applied to the Hi-R LC MLAs. Furthermore, the generated elemental images with different depths can be reconstructed at different CDPs by electrically controllable focal lengths of the Hi-R LC MLAs, and these 3D scenes can be realized the switchable display in the same II 3D display system, which can effectively present the optical performance of the II 3D display.

Spatially incongruent sounds affect visual localization in virtual environments.

Distance underestimations along the depth plane are widely found in virtual environments. However, past findings have shown that changes in the visual aspects of virtual reality settings do not lead to more accurate depth estimates. Therefore, we examined if nonvisual stimuli, namely, sounds, could serve as cues that affect observers' depth perception. Accordingly, we conducted two distance discrimination tasks to examine whether observers' depth localization is affected by a spatially incongruent sound. In Experiment 1, a spatially incongruent sound made a visual target appear farther away than a visual target presented with no sound only when a far-distance range (i.e., longer than 12 m) was introduced. Experiment 2 further indicated that the sound shifted visual localization only when audiovisual spatial disparity did not exceed 4°. Taken together, our findings suggest that the depth localization of a visual object in virtual reality can be altered by a spatially incongruent sound, and provide a potential approach that we can adopt a spatially incongruent sound as a cue to reduce the depth compression in VR.