This PDF file contains the front matter associated with SPIE Proceedings Volume 8648, including the Title Page, Copyright information, Table of Contents, Introduction, and Conference Committee listing.

The MPEG 3DV project is working on the next generation video encoding standard and in this process a call for proposal of encoding algorithms was issued. To evaluate these algorithm a large scale subjective test was performed involving Laboratories all over the world. For the participating Labs it was optional to administer a slightly modified Simulator Sickness Questionnaire (SSQ) from Kennedy et al (1993) before and after the test. Here we report the results from one Lab (Acreo) located in Sweden. The videos were shown on a 46 inch film pattern retarder 3D TV, where the viewers were using polarized passive eye-glasses to view the stereoscopic 3D video content. There were 68 viewers participating in this investigation in ages ranges from 16 to 72, with one third females. The questionnaire was filled in before and after the test, with a viewing time ranging between 30 min to about one and half hour, which is comparable to a feature length movie. The SSQ consists of 16 different symptoms that have been identified as important for indicating simulator sickness. When analyzing the individual symptoms it was found that Fatigue, Eye-strain, Difficulty Focusing and Difficulty Concentrating were significantly worse after than before. SSQ was also analyzed according to the model suggested by Kennedy et al (1993). All in all this investigation shows a statistically significant increase in symptoms after viewing 3D video especially related to visual or Oculomotor system.

Stereoscopic 3D is undoubtedly one of the most attractive content. It has been deployed intensively during the last decade through movies and games. Among the advantages of 3D are the strong involvement of viewers and the increased feeling of presence. However, the sanitary e ects that can be generated by 3D are still not precisely known. For example, visual fatigue and visual discomfort are among symptoms that an observer may feel. In this paper, we propose an investigation of visual fatigue generated by 3D video watching, with the help of eye-tracking. From one side, a questionnaire, with the most frequent symptoms linked with 3D, is used in order to measure their variation over time. From the other side, visual characteristics such as pupil diameter, eye movements ( xations and saccades) and eye blinking have been explored thanks to data provided by the eye-tracker. The statistical analysis showed an important link between blinking duration and number of saccades with visual fatigue while pupil diameter and xations are not precise enough and are highly dependent on content. Finally, time and content play an important role in the growth of visual fatigue due to 3D watching.

Stereoscopic photography became popular soon after the introduction of photographic processes by Daguerre and by Talbot in 1839. Stereoscopic images were most often viewed as side-by-side left- and right-eye image pairs, using viewers with prisms or mirrors. Superimposition of encoded image pairs was envisioned as early as the 1890s, and encoding by polarization first became practical in the 1930s with the introduction of polarizers in large sheet form. The use of polarizing filters enabled projection of stereoscopic image pairs and viewing of the projected image through complementary polarizing glasses. Further advances included the formation of images that were themselves polarizers, forming superimposed image pairs on a common carrier, the utilization of polarizing image dyes, the introduction of micropolarizers, and the utilization of liquid crystal polarizers.

We suggested a mathematical model describing optical phenomena at the surface of a lenticular array for autostereoscopic displays in 2-view system and simulated the angular distribution of light using Monte Carlo methods under actual design conditions. By comparing the results of our simulation model with those of the conventional simulation tool, we showed good agreement between our model and conventional tool in calculating the angular distribution of light. We also achieved simulation time of 60% faster than the conventional design tool. From the results, we think our optical model is very useful for designing optical parameters of autostereoscopic displays.

Lenticular lens based autostereoscopic 3D displays are finding many applications in digital signage and consumer electronics devices. A high quality 3D viewing experience requires the lenticular lens be properly aligned with the pixels on the display device so that each eye views the correct image. This work presents a simple and novel method for rapidly assessing the quality of a lenticular lens to be used in autostereoscopic displays. Errors in lenticular alignment across the entire display are easily observed with a simple test pattern where adjacent views are programmed to display different colors.

Crosstalk is one of the main stereoscopic display-related visual perceptual factors degrading image quality and causing visual discomfort. In this research the impact of high display contrast and high display luminance on the perception of crosstalk is investigated by using a custom-built high-dynamic range LCD (liquid-crystal display) in combination with a Wheatstone viewer. The displays’ opto-electrical response was characterized and the display calibrated, to independently vary luminance, contrast, and crosstalk (defined as (BW − BB) ⁄ (WW − BB)). The crosstalk visibility threshold was determined via a ‘one-up/two-down’ staircase method by fourteen participants for three different images that varied in luminance (125, 500, and 1,500 cd/m²) and contrast (1,000:1 and 2,500:1). Results show that an increase in luminance leads to a reduced crosstalk visibility threshold to a minimal value of 0.19% at 1,500 cd/m². The crosstalk visibility threshold was independent of the tested contrast levels, indicating that contrast levels above 100:1 do not affect crosstalk visibility thresholds. Important to note is that for displays with high contrast, the finite discrete levels of transmission in the LC-panel quantize the luminance levels, which propagates into and limits the accuracy of the crosstalk visibility threshold. In conclusion, by introducing OLEDs (high contrast), the system crosstalk will increase by definition, but visibility of crosstalk will not. By introducing high-dynamic range displays (high peak luminance), the crosstalk visibility threshold will be lower. As the absolute threshold levels of low-dynamic range displays are already very low (at or below 0.3%) this will result in little perceptual effect.

Liquid Crystal Shutter (LCS) 3D glasses (also known as 3D shutter glasses, active glasses, or just shutter glasses) are the selection device commonly used to view 3D stereoscopic content on time-sequential 3D displays. There are many available models of LCS glasses to choose from; although, unfortunately, there can be wide performance differences between these glasses. Ultimately the perceived quality of 3D image viewing depends on the display viewed as well as the performance of the shutter glasses used. The objective of this paper is to define key glasses performance parameters and to present a testing method to help quantify glasses and select between models available. Differences in shutter opening speed and open transparency (shuttering performance) are shown between tested models using the method presented here. Additional differences are noted for other performance parameters and features illustrating that there can be many product differences to consider that do not directly affect shuttering performance.

Subjective assessment of Quality of Experience in stereoscopic 3D requires new guidelines for the environmental setup as existing standards such as ITU-R BT.500 may no longer be appropriate. A first step is to perform cross-lab experiments in different viewing conditions on the same video sequences. Three international labs performed Absolute Category Rating studies on a freely available video database containing degradations that are mainly related to video quality degradations. Different conditions have been used in the labs: Passive polarized displays, active shutter displays, differences in viewing distance, the number of parallel viewers, and the voting device. Implicit variations were introduced due to the three different languages in Sweden, South Korea, and France. Although the obtained Mean Opinion Scores are comparable, slight differences occur in function of the video degradations and the viewing distance. An analysis on the statistical differences obtained between the MOS of the video sequences revealed that obtaining an equivalent number of differences may require more observers in some viewing conditions. It was also seen that the alignment of the meaning of the attributes used in Absolute Category Rating in different languages may be beneficial. Statistical analysis was performed showing influence of the viewing distance on votes and MOS results.

We propose a new method which compensates optical axes misalignment along with stereo camera zooming automatically. Optical axes misalignment is modeled as image sensor center translation from the lens center. To match the image points of left and right image, right camera translation and rotation equation is devised. By using these equations we can easily get the rig calibration value. Then, we can compensate optical axes misalignment by saving and applying this rig calibration information when we change the zoom of stereo camera. The explanation of the proposed method and devised equation will be provided. And experimental verification results using stereo camera rig will be presented in the paper.

The rendering of photorealistic rain has been previously studied for monoscopic viewing. We extend the monoscopic statistical rain models to simulate the behavior and distribution of falling rain for stereo viewing. Our goal is to be able to render in real-time frame rates. In this investigation we ignore the complex issues of scene illumination and concentrate on the parameters that produce a realistic rain distribution. Using the concept of retinal persistence we render a visible falling raindrop as a linear streak. To speed rendering we use pre-computed images of such rain streaks. Rain streak positions for the left- and right-eye views are created by generating random numbers that depend on the view volume of the scene. We permit interactive but controlled modification of rain parameters such as density and wind gusts. We compare our approach to the use of existing 2D-3D conversion methods. The results demonstrate that using commercial 2D-3D converters are not sufficient in producing realistic stereo rain effects. Future research will concentrate on including complex lighting interactions.

Multi-view three-dimensional television requires many views, which may be synthesized from two-dimensional images with accompanying pixel-wise depth information. This depth image, which typically consists of smooth areas and sharp transitions at object borders, must be consistent with the acquired scene in order for synthesized views to be of good quality. We have previously proposed a depth image coding scheme that preserves significant edges and encodes smooth areas between these. An objective evaluation considering the structural similarity (SSIM) index for synthesized views demonstrated an advantage to the proposed scheme over the High Efficiency Video Coding (HEVC) intra mode in certain cases. However, there were some discrepancies between the outcomes from the objective evaluation and from our visual inspection, which motivated this study of subjective tests. The test was conducted according to ITU-R BT.500-13 recommendation with Stimulus-comparison methods. The results from the subjective test showed that the proposed scheme performs slightly better than HEVC with statistical significance at majority of the tested bit rates for the given contents.

This paper presents a complete framework for capturing, processing and displaying the free viewpoint video on a large scale immersive light-field display. We present a combined hardware-software solution to visualize free viewpoint 3D video on a cinema-sized screen. The new glasses-free 3D projection technology can support larger audience than the existing autostereoscopic displays. We introduce and describe our new display system including optical and mechanical design considerations, the capturing system and render cluster for producing the 3D content, and the various software modules driving the system. The indigenous display is first of its kind, equipped with front-projection light-field HoloVizio technology, controlling up to 63 MP. It has all the advantages of previous light-field displays and in addition, allows a more flexible arrangement with a larger screen size, matching cinema or meeting room geometries, yet simpler to set-up. The software system makes it possible to show 3D applications in real-time, besides the natural content captured from dense camera arrangements as well as from sparse cameras covering a wider baseline. Our software system on the GPU accelerated render cluster, can also visualize pre-recorded Multi-view Video plus Depth (MVD4) videos on this light-field glasses-free cinema system, interpolating and extrapolating missing views.

State-of-the-art autostereoscopic displays are often limited in size, effective brightness, number of 3D viewing zones, and maximum 3D viewing distances, all of which are mandatory requirements for large-scale outdoor displays. Conventional autostereoscopic indoor concepts like lenticular lenses or parallax barriers cannot simply be adapted for these screens due to the inherent loss of effective resolution and brightness, which would reduce both image quality and sunlight readability. We have developed a modular autostereoscopic multi-view laser display concept with sunlight readable effective brightness, theoretically up to several thousand 3D viewing zones, and maximum 3D viewing distances of up to 60 meters. For proof-of-concept purposes a prototype display with two pixels was realized. Due to various manufacturing tolerances each individual pixel has slightly different optical properties, and hence the 3D image quality of the display has to be calculated stochastically. In this paper we present the corresponding stochastic model, we evaluate the simulation and measurement results of the prototype display, and we calculate the achievable autostereoscopic image quality to be expected for our concept.

A multi-view display module using microelectromechanical system (MEMS) projectors is proposed to realize ultra-large screen autostereoscopic displays. The module consists of an array of MEMS projectors, a vertical diffuser, and a lenticular lens. All MEMS projectors having different horizontal positions project images that are superimposed on the vertical diffuser. Each cylindrical lens constituting the lenticular lens generates multiple three-dimensional (3D) pixels at different horizontal positions near its focal plane. Because the 3D pixel is an image of a micro-mirror of the MEMS projector, the number of 3D pixels in each lens is equal to the number of MEMS projectors. Therefore, the horizontal resolution of the module can be increased using more projectors. By properly modulating lasers in the MEMS projector, the horizontal positions of dots constituting a projected image can be altered at different horizontal scan lines. By increasing the number of scan lines corresponding to one 3D pixel, the number of views can be increased. Because the module has a frameless screen, a number of modules can be arranged two-dimensionally to obtain a large screen. The prototype module was constructed using four MEMS projectors. The module had a 3D resolution of 160 × 120 and provided 64 views. The screen size was 14.4 inches.

A Multi-Layer Display (MLD) consists of two or more imaging planes separated by physical depth where the depth is a key component in creating a glasses-free 3D effect. Its core benefits include being viewable from multiple angles, having full panel resolution for 3D effects with no side effects of nausea or eye-strain. However, typically content must be designed for its optical configuration in foreground and background image pairs. A process was designed to give a consistent 3D effect in a 2-layer MLD from existing stereo video content in real-time. Optimizations to stereo matching algorithms that generate depth maps in real-time were specifically tailored for the optical characteristics and image processing algorithms of a MLD. The end-to-end process included improvements to the Hierarchical Belief Propagation (HBP) stereo matching algorithm, improvements to optical flow and temporal consistency. Imaging algorithms designed for the optical characteristics of a MLD provided some visual compensation for depth map inaccuracies. The result can be demonstrated in a PC environment, displayed on a 22” MLD, used in the casino slot market, with 8mm of panel seperation. Prior to this development, stereo content had not been used to achieve a depth-based 3D effect on a MLD in real-time

The Switching Dual Layer Display with a Dynamic LCD Mask is a multi-layer three-dimensional display capable of independent layer content and occlusion between layers. The display consists of a spaced stack of transparent LCD panel and LCD monitor synchronously switching between content, mask, and backlight states. In the first state, the back layer displays white and the front layer displays its content. The back layer acts as a backlight for the front layer, making the front layer content visible. In the second state, the back layer displays content and the front layer displays a mask. The back layer content is occluded by the front layer mask. Rapidly alternating between states, the viewer perceives opaque high-contrast foreground content occluding the background. Besides multi-plane imagery, depth-blending using opacity allows for smooth 3D volumetric imagery at the cost of reduced field of view. The addition of a relay mirror and a phantom mask allows solid appearing front layer content to optically float above and occlude the back layer. The floating front layer is non-physical, so both display layers are accessible for interactive applications.

Slice-stacking displays are particular kind of volumetric displays, which present a large number of volumetric elements to viewers. These elements are typically packed in a 2D array, called “slice” or “depth plane”, which sequentially fills a 3D volume. Recently, multilayer displays have been proposed and implemented by several researchers. Unlike slice-stacking displays, multilayer displays can present multiple slices at the same time. Furthermore, slices in multilayer displays always remain stationary, i.e., there are no sliding or rotating slices, which are often found in slice-stacking displays. In this paper, we consider a mixture of slice-stacking and multilayer displays. More specifically, we present a new design of multilayer displays, where slices will move periodically and translationally changing the relative distances from the viewers. By allowing the slices to move around, we can, in effect, increase the number of layers in the displays. This results in finer resolution of the light field observable from the displays, and improvement in the precision of the accommodation cues. An implementation of the new design using a beam splitter and a single moving mirror is also presented along with a few experimental results.

We can successfully solve the problem in DFD display that the maximum depth difference of front and rear planes is limited because depth fusing from front and rear images to one 3-D image becomes impossible. The range of continuously perceived depth was estimated as depth difference of front and rear planes increases. When the distance was large enough, perceived depth was near front plane at 0~40 % of rear luminance and near rear plane at 60~100 % of rear luminance. This maximum depth range can be successfully enlarged by spatial-frequency modulation of front and rear images. The change of perceived depth dependence was evaluated when high frequency component of front and rear images is cut off using Fourier Transformation at the distance between front and rear plane of 5 and 10 cm (4.9 and 9.4 minute of arc). When high frequency component does not cut off enough at the distance of 5 cm, perceived depth was separated to near front plane and near rear plane. However, when the images are blurred enough by cutting high frequency component, the perceived depth has a linear dependency on luminance ratio. When the images are not blurred at the distance of 10 cm, perceived depth is separated to near front plane at 0~30% of rear luminance, near rear plane at 80~100 % and near midpoint at 40~70 %. However, when the images are blurred enough, perceived depth successfully has a linear dependency on luminance ratio.

As stereoscopic displays have spread, it is important to know what really causes the visual fatigue and discomfort and what happens in the visual system in the brain behind the retina while viewing stereoscopic 3D images on the displays. In this study, functional magnetic resonance imaging (fMRI) was used for the objective measurement to assess the human brain regions involved in the processing of the stereoscopic stimuli with excessive disparities. Based on the subjective measurement results, we selected two subsets of comfort videos and discomfort videos in our dataset. Then, a fMRI experiment was conducted with the subsets of comfort and discomfort videos in order to identify which brain regions activated while viewing the discomfort videos in a stereoscopic display. We found that, when viewing a stereoscopic display, the right middle frontal gyrus, the right inferior frontal gyrus, the right intraparietal lobule, the right middle temporal gyrus, and the bilateral cuneus were significantly activated during the processing of excessive disparities, compared to those of small disparities (< 1 degree).

In the conventional temporally interlaced S3D protocol, red, green, and blue are presented simultaneously to one eye and then to the other eye. Thus, images are presented in alternating fashion to the two eyes. Moving objects presented with this protocol are often perceived at incorrect depth relative to stationary parts of the scene. We implemented a colorinterlaced protocol that could in principle minimize or even eliminate such depth distortions. In this protocol, green is presented to one eye and red and blue to the other eye at the same time. Then red and blue are presented to the first eye and green to the second. Using a stereoscope, we emulated the color-interlaced protocol and measured the magnitude of depth distortions as a function of object speed. The results showed that color interlacing yields smaller depth distortions than temporal interlacing in most cases and never yields larger distortions. Indeed, when color interlacing produces no brightness change within sub-frames, the distortions are eliminated altogether. The results also show that the visual system’s calculation of depth from disparity is based on luminance, not chromatic signals. In conclusion, color interlacing provides great potential for improved stereo presentation.

Purpose: The study evaluated the accuracy of depth perception afforded by static and dynamic stereoscopic three-dimensional (S3D) images with proportional (scaled to disparity) and constant size cues. Methods: Sixty adult participants, 18 to 40 years (mean, 24.8 years), with good binocular vision participated in the study. For static S3D trials, participants were asked to indicate the depth of stationary S3D images rendered with 36, 48 and 60 pixels of crossed disparity, and with either proportional or a constant size. For dynamic S3D trials, participants were asked to indicate the time when S3D images, moving at 27, 32 and 40 pixels/sec, matched the depth of a reference image which was presented with 36, 48 and 60 pixels of crossed image disparity. Results: Results show that viewers perceived S3D images as being closer than would be predicted by the magnitude of image disparity, and correspondingly they overestimated the depth in moving S3D images. The resultant depth perception and estimate of motion speed were more accurate for conditions with proportional and larger image size, slower motion-in-depth and larger image disparity. Conclusion: These findings possibly explain why effects such as looming are over stimulating in S3D viewing. To increase the accuracy of depth perception, S3D content should match image size to its disparity level, utilize larger depth separation (without inducing excessive discomfort) and render slower motion in depth.

The fundamental visual property that drives 3D stereoscopic technology is the compelling qualitative experience of tangible solid objects, immersive space and realism that is lacking in conventional 2D displays. This qualitative visual phenomenon, referred to as ‘stereopsis’, is widely assumed to be a by-product of binocular vision. However, its underlying cause, variation and functional role remain largely unexplained. In this theoretical paper I briefly present an alternative hypothesis that stereopsis is not a phenomenon restricted to binocular vision, but a more basic qualitative visual property related to the perception of egocentric distance and scale. I review recent empirical evidence showing that stereopsis is not simply a product of binocular disparities or the mere perception of “more depth”. The theory and results imply critical distinctions between the qualitative experience of stereopsis and the quantitative perception of 3D structure. I describe how this alternative view has important implications for the perception of scale and realism in both conventional and stereoscopic display systems; e.g., perception of miniaturization (puppet-theater effect) and gigantism.

We propose a new optical configuration to form aerial 3D LEDs by use of retroreflective sheeting. The proposed configuration is composed of LEDs, a half mirror, and retroreflective sheeting. A half of the LED lights are reflected by the half mirror and impinge on the retroreflective sheeting. The retroreflective sheeting reflects the lights back to their sources. On the way to the sources, a half of the lights transmit through the half mirror and form the real images of LEDs. Although less than 25 % of the output lights are contributed to the aerial image, recent LED panels have enough luminance to enjoy the aerial image with a quarter of the luminance. We have made a prototype of the proposed aerial LED display. An aerial image of the LED panel has been successfully formed in free space. Its viewing angle was significantly improved compared to the aerial display by use of crossed mirrors, which limit the viewing angle by aperture size and height of mirror walls. The viewing angle in the proposed configuration is mainly limited by the size of the retroreflective sheeting. Furthermore, by using LEDs in different depths, we realized an aerial 3D display in free space.

A holographic rendering algorithm using a layer-based structure with angular tiling supports view-dependent shading and accommodation cues. This approach also has the advantages of rapid computation speed and visual reduction of layer gap artefacts compared to other approaches. Holograms rendered with this algorithm are displayed using an SLM to demonstrate view-dependent shading and occlusion.

We present a method to synthesize a hologram of real-existing three-dimensional object under incoherent illumination. The proposed method uses an integral imaging camera which has a microlens array on the image sensor plane of usual two-dimensional camera. The three-dimensional object is captured by the integral imaging camera under regular incoherent illumination. The captured image is then processed to generate a number of different orthographic view images of the three-dimensional object. The synthetic aperture technique is also used to reduce the effective f-number of the microlens array. The proposed idea is verified by the experimental result.

With 3D televisions expected to reach 50% home saturation as early as 2016, understanding the psychological mechanisms underlying the user response to 3D technology is critical for content providers, educators and academics. Unfortunately, research examining the effects of 3D technology has not kept pace with the technology’s rapid adoption, resulting in large-scale use of a technology about which very little is actually known. Recognizing this need for new research, we conducted a series of studies measuring and comparing many of the variables and processes underlying both 2D and 3D media experiences. In our first study, we found narratives within primetime dramas had the power to shift viewer attitudes in both 2D and 3D settings. However, we found no difference in persuasive power between 2D and 3D content. We contend this lack of effect was the result of poor conversion quality and the unique demands of 3D production. In our second study, we found 3D technology significantly increased enjoyment when viewing sports content, yet offered no added enjoyment when viewing a movie trailer. The enhanced enjoyment of the sports content was shown to be the result of heightened emotional arousal and attention in the 3D condition. We believe the lack of effect found for the movie trailer may be genre-related. In our final study, we found 3D technology significantly enhanced enjoyment of two video games from different genres. The added enjoyment was found to be the result of an increased sense of presence.

We create a system for semi-automatically converting unconstrained 2D images and videos into stereoscopic 3D. Current efforts are done automatically or manually by rotoscopers. The former prohibits user intervention, or error correction, while the latter is time consuming, requiring a large staff. Semi-automatic mixes the two, allowing for faster and accurate conversion, while decreasing time to release 3D content. User-defined strokes for the image, or over several keyframes, corresponding to a rough estimate of the scene depths are defined. After, the rest of the depths are found, creating depth maps to generate stereoscopic 3D content, and Depth Image Based Rendering is employed to generate the artificial views. Here, depth map estimation can be considered as a multi-label segmentation problem, where each class is a depth value. Optionally, for video, only the first frame can be labelled, and the strokes are propagated using a modified robust tracking algorithm. Our work combines the merits of two respected segmentation algorithms: Graph Cuts and Random Walks. The diffusion of depths from Random Walks, combined with the edge preserving properties from Graph Cuts is employed to create the best results possible. Results demonstrate good quality stereoscopic images and videos with minimal effort.

Disparity estimation has been extensively investigated in recent years. Though several algorithms have been reported to achieve excellent performance on the Middlebury website, few of them reach a satisfying balance between accuracy and efficiency, and few of them consider the problem of temporal coherence. In this paper, we introduce a novel disparity estimation approach, which improves the accuracy for static images and the temporal coherence for videos. For static images, the proposed approach is inspired by the adaptive support weight method proposed by Yoon et al. and the dual-cross-bilateral grid introduced by Richardt et al. Principal component analysis (PCA) is used to reduce the color dimensionality in the cost aggregation step. This simple, but efficient technique helps the proposed method to be comparable to the best local algorithms on the Middlebury website, while still allowing real-time implementation. A computationally efficient method for temporally consistent behavior is also proposed. Moreover, in the user evaluation experiment, the proposed temporal approach achieves the best overall user experience among the selected comparison algorithms.

To ensure an immersive, yet comfortable experience, significant work is required during post-production to adapt the stereoscopic 3D (S3D) content to the targeted display and its environment. On the one hand, the content needs to be reconverged using horizontal image translation (HIT) so as to harmonize the depth across the shots. On the other hand, to prevent edge violation, specific re-convergence is required and depending on the viewing conditions floating windows need to be positioned. In order to simplify this time-consuming work we propose a depth grading tool that automatically adapts S3D content to digital cinema or home viewing environments. Based on a disparity map, a stereo point of interest in each shot is automatically evaluated. This point of interest is used for depth matching, i.e. to position the objects of interest of consecutive shots in a same plane so as to reduce visual fatigue. The tool adapts the re-convergence to avoid edge-violation, hyper-convergence and hyper-divergence. Floating windows are also automatically positioned. The method has been tested on various types of S3D content, and the results have been validated by a stereographer.

The great success of the three-dimensional (3D) digital cinema industry has opened up a new era of 3D content services. While we have witnessed a rapid surge of stereoscopic 3D services, the issue of viewing safety remains a possible obstacle to the widespread deployment of such services. In this paper, we propose a novel disparity remapping method to reduce the visual discomfort induced by fast change in disparity. The proposed remapping approach selectively adjusts the disparities of the discomfort regions where the fast change in disparity occurs. To this purpose, the proposed approach detects visual importance regions in a stereoscopic 3D video, which may have dominant influence on visual comfort in video frames, and then locally adjust the disparity by taking into account the disparity changes in the visual importance regions. The experimental results demonstrate that the proposed approach to adjust local problematic regions can improve visual comfort while preserving naturalness of the scene.

The authors have analyzed binocular disparity included in stereoscopic (3D) images from the perspective of producing depth sensation. This paper described the disparity analysis conducted by the authors for well-known 3D movies. Two types of disparity analysis were performed; full-length analysis of four 3D movies and analysis of emotional scenes from them. This paper reports an overview of the authors’ approaches and the results obtained from their analysis.

Creating and processing stereoscopic video imposes additional quality requirements related to view synchronization. In this work we propose a set of algorithms for detecting typical stereoscopic-video problems, which appear owing to imprecise setup of capture equipment or incorrect postprocessing. We developed a methodology for analyzing the quality of S3D motion pictures and for revealing their most problematic scenes. We then processed 10 modern stereo films, including Avatar, Resident Evil: Afterlife and Hugo, and analyzed changes in S3D-film quality over the years. This work presents real examples of common artifacts (color and sharpness mismatch, vertical disparity and excessive horizontal disparity) in the motion pictures we processed, as well as possible solutions for each problem. Our results enable improved quality assessment during the filming and postproduction stages.

An important aspect in the quality of the three-dimensional perception produced by any stereoscopic production is the precision of its left/right image registration. Tests show that pixel-precise stereoscopic image registration and convergence improves viewer’s three-dimensional perception by reducing viewing discomfort.¹ Current software tools generally rely on techniques like 50/50 (onionskin) for camera system and left/right post-production image alignment. Recent stereoscopic software toolset beta tests indicate that these alignment techniques make pixel-accurate registration difficult to achieve. A review of 70 stereo pairs sampled from seven stereoscopic feature films, released from 2009 to 2012, revealed that the majority lacked a precise point of convergence. Extended viewing of inaccurate stereoscopic convergence leads to viewer fatigue and discomfort.

This paper describes the development, measurement, computation, and initial testing of a metric of antialiasing sufficiency for stereoscopic display systems. A summary is provided of two previous evaluations that demonstrated stereoscopic disparity thresholds in the range of 3 to 10 arcsec are attainable using electronic displays with a pixel pitch as coarse as 2.5 arcmin, however, only if sufficient antialiasing is performed. Equations are provided that describe the critical level of antialiasing required as a function of pixel pitch. The proposed metric uses a radial test pattern that can be photographed from the user eyepoint using a hand held consumer color camera. Several candidate unitary metrics that quantify the spatial sampling noise in the measured test pattern were tested. The correlation obtained between the best candidate metric and the stereoscopic disparity threshold model from our previous paper was R² = 0.994. The standard deviation of repeated measurements with a hand held camera was less than 0.5% of the range of the metric, indicating the metric is capable of discriminating fine differences in sampling noise. The proposed method is display technology independent and requires no knowledge of the display pixel structure or how the antialiasing is implemented.

We report on a new game design where the goal is to make the stereoscopic depth cue sufficiently critical to success that game play should become impossible without using a stereoscopic 3D (S3D) display and, at the same time, we investigate whether S3D game play is affected by screen size. Before we detail our new game design we review previously unreported results from our stereoscopic game research over the last ten years at the Durham Visualisation Laboratory. This demonstrates that game players can achieve significantly higher scores using S3D displays when depth judgements are an integral part of the game. Method: We design a game where almost all depth cues, apart from the binocular cue, are removed. The aim of the game is to steer a spaceship through a series of oncoming hoops where the viewpoint of the game player is from above, with the hoops moving right to left across the screen towards the spaceship, to play the game it is essential to make decisive depth judgments to steer the spaceship through each oncoming hoop. To confound these judgements we design altered depth cues, for example perspective is reduced as a cue by varying the hoop's depth, radius and cross-sectional size. Results: Players were screened for stereoscopic vision, given a short practice session, and then played the game in both 2D and S3D modes on a seventeen inch desktop display, on average participants achieved a more than three times higher score in S3D than they achieved in 2D. The same experiment was repeated using a four metre S3D projection screen and similar results were found. Conclusions: Our conclusion is that games that use the binocular depth cue in decisive game judgements can benefit significantly from using an S3D display. Based on both our current and previous results we additionally conclude that display size, from cell-phone, to desktop, to projection display does not adversely affect player performance.

The floating window technique is commonly employed by stereoscopic 3D filmmakers to reduce the effects of window violations by masking out portions of the screen that contain visual information that doesn’t exist in one of the views. Although widely adopted in the film industry, and despite its potential benefits, the technique has not been adopted by video game developers to the same extent possibly because of the lack of understanding of how the floating window can be utilized in such an interactive medium. Here, we describe a quantitative study that investigates how the floating window technique affects users’ depth perception in a simple game-like environment. Our goal is to determine how various stereoscopic 3D parameters such as the existence, shape, and size of the floating window affect the user experience and to devise a set of guidelines for game developers wishing to develop stereoscopic 3D content. Providing game designers with quantitative knowledge of how these parameters can affect user experience is invaluable when choosing to design interactive stereoscopic 3D content.

Stereoscopic 3D (S3D) content in games, film and other audio-visual media has been steadily increasing over the past number of years. However, there are still open, fundamental questions regarding its implementation, particularly as it relates to a multi-modal experience that involves sound and haptics. Research has shown that sound has considerable impact on our perception of 2D phenomena, but very little research has considered how sound may influence stereoscopic 3D. Here we present the results of an experiment that examined the effects of 5.1 surround sound (5.1) and stereo loudspeaker setups on depth perception in relation to S3D imagery within a video game environment. Our aim was to answer the question: “can 5.1 surround sound enhance the participant’s perception of depth in the stereoscopic field when compared to traditional stereo sound presentations?” In addition, our study examined how the presence or absence of Doppler frequency shift and frequency fall-off audio effects can also influence depth judgment under these conditions. Results suggest that 5.1 surround sound presentations enhance the apparent depth of stereoscopic imagery when compared to stereo presentations. Results also suggest that the addition of audio effects such as Doppler shift and frequency fall-off filters can influence the apparent depth of S3D objects.

In this paper we undertook perceptual experiments to determine the allowed differences in depth between audio and visual stimuli in stereoscopic-3D environments while being perceived as congruent. We also investigated whether the nature of the environment and stimuli affects the perception of congruence. This was achieved by creating an audio-visual environment consisting of a photorealistic visual environment captured by a camera under orthostereoscopic conditions and a virtual audio environment generated by measuring the acoustic properties of the real environment. The visual environment consisted of a room with a loudspeaker or person forming the visual stimulus and was presented to the viewer using a passive stereoscopic display. Pink noise samples and female speech were used as audio stimuli which were presented over headphones using binaural renderings. The stimuli were generated at different depths from the viewer and the viewer was asked to determine whether the audio stimulus was nearer, further away or at the same depth as the visual stimulus. From our experiments it is shown that there is a significant range of depth differences for which audio and visual stimuli are perceived as congruent. Furthermore, this range increases as the depth of the visual stimulus increases.

Given the popularity of 3D film, content developers have been creating customizable stereoscopic 3D experiences for the user to enjoy at home. Stereoscopic 3D game developers often offer a ‘white box’ approach whereby far too many controls and settings are exposed to the average consumer who may have little knowledge or interest to correctly adjust these settings. Improper settings can lead to users being uncomfortable or unimpressed with their own user-defined stereoscopic 3D experience. We have begun investigating interactive approaches to in-game adjustment of the various stereoscopic 3D parameters to reduce the reliance on the user doing so and thefore creating a more pleasurable stereoscopic 3D experience. In this paper, we describe a preliminary technique for interactively calibrating the various stereoscopic 3D parameters and we compare this interface with the typical slider-based control interface game developers utilize in commercial S3D games. Inspired by standard testing methodologies experienced at an optometrist, we’ve created a split-screen game with the same stereoscopic 3D game running in both screens, but with different interaxial distances. We expect that the interactive nature of the calibration will impact the final game experience providing us with an indication of whether in-game, interactive, S3D parameter calibration is a mechanism that game developers should adopt.

Super multi-view (SMV) displays have been developed to solve the vergence-accommodation conflict that causes visual fatigue with conventional three-dimensional (3D) displays. An ideal SMV display should generate dense viewpoints in both the horizontal and vertical directions. However, the SMV displays developed so far generate dense viewpoints only in the horizontal direction because of the system complexity required for full-parallax systems. Therefore, the accommodation responses to SMV displays were measured mostly for 3D images that contain distinctive vertical edges. In this study, we investigated the influences of the edge properties contained in 3D images upon the accommodation responses evoked by horizontal-parallax-only (HPO) SMV displays. We used the recently developed reduced-view SMV display, whose interval of viewpoints was 2.6 mm, for the accommodation measurements. Two test images were used: a “Maltese cross”, which contains several sharp edges, and a standard test image of “Lenna”, which contains various types of edges. We found that the HPO–SMV display still evoked the accommodation responses when “Lenna” was displayed. There were two types of accommodation responses for “Lenna”; the amounts of the evoked accommodation were smaller than or similar to those for the “Maltese cross”, depending on where the viewers gazed.

Conventional stereoscopic displays render 3D scenes from a single pair of 2D images at a fixed distance to the viewer. Therefore conventional stereoscopic displays force an unnatural decoupling of the accommodation and convergence cues, which may contribute to various visual artifacts and have significant effects on depth perception accuracy. In this paper, we present the design and implementation of an optical see-through depth-fused multi-focal-plane binocular stereoscopic display. The prototype is capable of rendering nearly-correct focus cues for a large volume of 3D objects extending into a depth range from 0 to 3 diopters and a diagonal field of view of 40 degrees at a flicker-free speed. With the optimized optical system including a freeform prism eyepiece and a see-through compensator, the prototype system demonstrates high image quality while having minimal degradation to the see-through view.

Game developers have yet to embrace and explore the interactive stereoscopic 3D medium. They typically view stereoscopy as a separate mode that can be disabled throughout the design process and rarely develop game mechanics that take advantage of the stereoscopic 3D medium. What if we designed games to be S3D-specific and viewed traditional 2D viewing as a separate mode that can be disabled? The design choices made throughout such a process may yield interesting and compelling results. Furthermore, we believe that interaction within a stereoscopic 3D environment is more important than the visual experience itself and therefore, further exploration is needed to take into account the interactive affordances presented by stereoscopic 3D displays. Stereoscopic 3D displays allow players to perceive objects at different depths, thus we hypothesize that designing a core mechanic to take advantage of this viewing paradigm will create compelling content. In this paper, we describe Z-Fighter a game that we have developed that requires the player to interact directly along the stereoscopic 3D depth axis. We also outline an experiment conducted to investigate the performance, perception, and enjoyment of this game in stereoscopic 3D vs. traditional 2D viewing.

Purpose: Stereoscopic three-dimensional (S3D) viewing enhances depth perception of two-dimensional (2D) images. The present study measured viewer’s ability to discern depth differences and depth change afforded by image disparities presented on an S3D display. Methods: Sixty adults (age 24.8 +/- 3.4 years, 28% male) with binocular acuity better than 20/25 and stereoacuity better than 60 arcsec viewed test images presented on a 55” 3D TV (1920 x 1080 pixels) at 2.4m distance. In each trial, three of the four circles in the test image were with the same crossed baseline disparity of 12, 24, 36, 48, and 60 pixels, and the other (target) circle with added crossed disparity (delta disparity) of 2, 4, and 6 pixels. A subsequent change in delta disparity of the target circle (+/-2 pixels) was presented. Participant's response time and accuracy for identifying the target circle and its direction of depth change, as well as their vergence eye position, was recorded. Results: Larger baseline disparity resulted in lower accuracy and longer response times in identifying the target circle. The change of delta disparity was more accurately discerned when the delta disparity was larger and the change resulted in increased crossed disparity. Direction of vergence change and the final vergence error reflected an averaging of screen and image depths. Conclusion: S3D-induced depth difference and motion in depth is more easily discerned with smaller crossed disparity for background objects (< 43.3 arcmin or 48 pixels) and larger separation between image disparities (>3.6 arcmin or 4 pixels).

As 3D technology spreads, 3D imagery is being viewed in more diverse situations. Already, televisions and mobile devices are able to present 3D images, in addition to the cinema. Of these, mobile devices have the unique characteristic of requiring users to view images at close range. The authors conduct an experiment to examine the relationship between visual comfort and an individual observer's near phoria and interpupillary distance. The results show that observers with a higher degree of exophoria tend to prefer stereo images behind the screen because they cause less visual fatigue. On the other hand, observers with little exophoria or esophoria tend to prefer stereo images in front of the screen. Further, the results also show that observers with a greater interpupillary distance tend to prefer stereo images behind the screen. These findings suggest that the ability to adjust the depth of stereo images in advance based on personal phoria or interpupillary distance might help users achieve comfortable stereo viewing on mobile devices.

Introduction: The presentation of surgical contents to undergraduate medical students can be challenging, as the surgical approach is often different from the anatomist’s perspective that is reproduced in textbooks. Although there are many options to record endoscopic, microscopic as well as “open” surgical procedures, presentation of contents still can be costly and entail a loss in picture quality including depth impression. Material and methods: We presented seven stereoscopic clips of 30 seconds to minute and 20 seconds each to 64 medical students (43 female / 21 male) as part of the “sensory organs” course module in 4th year; using one 55” LCD 3D screen with line-alternating, circular polarization. Students were asked for their subjective viewing impression and about their opinion on the usefulness of 3D presentations in medical lectures. Results: 63% of students returned their questionnaires completed. The main results (multiple answers allowed) were: 70% noted that 3D presentations made complex anatomy easier to comprehend from an unknown perspective, 48% would feel better motivated to learn surgical procedures, and 38% would generally prefer a 3D lecture to a 2D lecture, while 23% would not see any advantage of 3D presentations whatsoever. Conclusion: While the screen size compared to audience size was far from ideal, it gave medical students, who had not been exposed to surgical procedures in the operating theatre yet, an impression of general approach to microsurgery and how the choice of surgical approach in relation of vital structures can minimize trauma and unwanted effects to the patient. The availability of larger screens, however, may necessitate changes in production of 3D material from the microscope camera onward.

Immersive stereoscopic imaging requires sharp wide field images, special software, and high resolution displays. Examples of some successful image capture, splicing, viewing, hosting, and posting techniques used in digital stereoscopic panoramic photography are given. Image capture uses camera movements that approximate natural eye positions reasonably well by using manual or motorized gimbal mounted systems designed for the purpose. Processing requires seamlessly stitching dozens or hundreds of images into left and right panoramas. Creating stereoscopic images over 50 mega pixels benefits from programmable motorized camera mounts. The 2 gig limit of TIFFs is often exceeded and requires the use of GigaPan.org hosting technologies. Gigapixel stereoscopic images are viewed as a single whole while many small files are quickly uploaded to improve the sharpness of the areas viewed and may be seen at 3d-360.com. Immersive stereo contents, active scrolling and deep zoom capabilities take stereoscopic photography from snapshots into the realm of immersive virtual presence when combined with modern web and display technologies. Scientific, artistic, and commercial applications can make effective use of existing stereoscopic displays systems by using these extended capabilities.

We are developing a bubble-projection three-dimensional display (BP3DD) that will generate reflecting points in a liquid. This type of display is currently nonexistent. In this paper, we examined bubble generation in water and ethanol to find the most suitable conditions for the BP3DD. Experimental results showed that when the water pressure was lower, the required energy to generate bubbles decreased linearly. The required energy should be lowest when the water pressure is similar to the water vapor pressure. When bubbles were generated in ethanol solution, the required energy was lower than that required in water. The minimum required energy to generate bubbles was approximately 2 J. We can conclude that the required energy to generate bubbles is dependent on the physical properties of the liquid and its pressure.

Integral Imaging is a technique to obtain true color 3D images that can provide full and continuous motion parallax for several viewers. The depth of field of these systems is mainly limited by the numerical aperture of each lenslet of the microlens array. A digital method has been developed to increase the depth of field of Integral Imaging systems in the reconstruction stage. By means of the disparity map of each elemental image, it is possible to classify the objects of the scene according to their distance from the microlenses and apply a selective deconvolution for each depth of the scene. Topographical reconstructions with enhanced depth of field of a 3D scene are presented to support our proposal.

We propose a polarization dependent double reflection technique to enhance the depth range of a multi-layered threedimensional display. The multi-layered three-dimensional display presents three-dimensional images by reconstructing light ray field using a stack of several liquid crystal display panels. The proposed system consists of a multi-layered display, a half-mirror, a polarization control panel, two pieces of quarter-wave-plates, a reflective polarizer and a backlight. The rays from the multi-layered display are controlled while passing through the polarization control panel to be either of two orthogonal polarization states. In one polarization state, the rays pass through the two quarter-waveplates, the half-mirror and the reflective polarizer directly. In another polarization state, the rays once reflect back from the reflective polarizer to the half-mirror before finally going out through the reflective polarizer, increasing apparent distance from the observer to the multi-layered display. By altering two polarization modes sequentially, the depth range of the multi-layered display can be doubled.

Depth estimation is an essential task for natural 3D image generation. In this paper, we estimate an accurate depth map from stereoscopic images captured by arbitrary camera arrays. Usually the depth information is estimated by stereo matching from two input images that are obtained by the parallel camera array. Recently the arc camera array has been widely employed to produce 3D movies. However, in the convergent camera array, it is difficult to apply image rectification by matching correspondence points due to serious image distortion. In this work, we estimate depth data without using image rectification. Once we define a potential energy function for depth detection based on spatial consistency, the energy optimization process identifies mismatching depth pixels. A reasonable depth value is assigned to each mismatched pixel using distance and intensity differences between the mismatched pixel and its neighbors. In addition, we improve temporal consistency and reduce visual discomfort. Experimental results demonstrate that our proposed method provides more accurate depth values than other methods based on image rectification.

The 3D video quality is of highest importance for the adoption of a new technology from a user’s point of view. In this paper we evaluated the impact of coding artefacts on stereoscopic 3D video quality by making use of several existing full reference 2D objective metrics. We analyzed the performance of objective metrics by comparing to the results of subjective experiment. The results show that pixel based Visual Information Fidelity metrics fits subjective data the best. The 2D stereoscopic video quality seems to have dominant impact on the coding artefacts impaired stereoscopic videos.

Marvel’s THE AVENGERS was the third film Stereo D collaborated on with Marvel; it was a summation of our artistic development of what Digitally Created 3D and Stereo D’s artists and toolsets affords Marvel’s filmmakers; the ability to shape stereographic space to support the film and story, in a way that balances human perception and live photography. We took our artistic lead from the cinematic intentions of Marvel, the Director Joss Whedon, and Director of Photography Seamus McGarvey. In the digital creation of a 3D film from a 2D image capture, recommendations to the filmmakers cinematic techniques are offered by Stereo D at each step from pre-production onwards, through set, into post. As the footage arrives at our facility we respond in depth to the cinematic qualities of the imagery in context of the edit and story, with the guidance of the Directors and Studio, creating stereoscopic imagery. Our involvement in The Avengers was early in production, after reading the script we had the opportunity and honor to meet and work with the Director Joss Whedon, and DP Seamus McGarvey on set, and into post. We presented what is obvious to such great filmmakers in the ways of cinematic techniques as they related to the standard depth cues and story points we would use to evaluate depth for their film. Our hope was any cinematic habits that supported better 3D would be emphasized. In searching for a 3D statement for the studio and filmmakers we arrived at a stereographic style that allowed for comfort and maximum visual engagement to the viewer.

We propose an omnidirectional three-dimensional (3D) display system. This is a tool for communication around a 3D image between a small number of people. This 3D display consists of multiple basic 3D display units. Each basic 3D display unit consists of a projector and a curved HOE lens array sheets. The projector is located on the center of the curvature of the curved HOE lens array, and it projects light rays on the curved HOE lens array with ascending vertical angle 45-degree. Projected light rays are reflected and reconstruct the 3D image over the center of the curvature by the curved HOE lens array. In this method, even though the viewing angle of each HOE lens does not increase, the viewing angle of the 3D image by the curved HOE lens array can be increased. To verify the effectiveness of the proposed method, we constructed the prototype basic 3D unit. The viewing angle of a 3D image was 44-degree. Therefore, the viewing angle was increased by the proposed method compared to 17-degree by the conventional flat HOE lens array. This paper describes the principle of proposed 3D display system, and also describes the experimental results.

We propose to use a rotational optical scanning method for a volumetric three-dimensional (3D) display based on optical scanning of an inclined image plane. The image plane was moved by an image rotator and formed by a dihedral corner reflector array, which is a distortion free imaging element forming a real image at a plane-symmetrical position. A stack of the moved and formed image planes of a two-dimensional (2D) display created displayable space of a 3D image, which satisfies all the criteria of stereoscopic vision and can be seen by the naked eyes. The image rotator used in this study was constructed of prism sheets and planar mirrors. The image plane was moved rotationally by rotating the proposed image rotator on an axis parallel to the mirror plane. Enlargement of the displayable space is achieved by just extending the distance between the 2D display and the mirror, so that the proposed scanning method is effective to display a large 3D image for our volumetric display system. We made a prototype display and observed a moved image plane to measure the displayable space of the 3D image. The size of the displayable space was 1200 [cm3] that is approximately six times as large as our previous display using a translational optical scanning method.

Super multi-view display provides smooth motion parallax without special glasses, and it is expected that the observer is free from the visual fatigue caused by the accommodation-vergence conflict. However, there is a problem that a huge number of pixels are required on a display device such as liquid crystal display panel because high-density rays are required for good quality images and each ray needs corresponding pixel. In this paper, we propose a new three-dimensional display based on lenticular method to reduce the required number of pixels. The rays are shot out to only around observer’s pupils. To do this, the lenticular lens of which viewing zone angle is narrowed is used and the lenticular lens is illuminated by parallel light made by cylindrical lenses and LEDs to suppress sidelobes. The direction of the parallel light is changed to follow the observer’s eye. We designed the optics system and confirmed the availability of the proposed method by computer simulation. Moreover, we constructed a prototype display and demonstrated. As a result, the parallax pitch and the viewing zone nearly equaled to designed values when the display was observed from the front, but these values were increased with the increasing viewing angle. The result of analysis, the reason why the parallax pitch and the viewing zone were expanded is thought as the curvature of field of the lenticular lens.

In this paper, we propose an autostereoscopic display system based on active anaglyph parallax barrier, which provides four viewpoints in full resolution. This system is realized with 120Hz displays and requires no special optical devices. With the four viewpoints achieved, a smoother multi-view experience can be achieved when involving head tracking. In addition, this system can be used for two-viewpoint autostereoscopy that allows more freedom of movement than that of a conventional parallax barrier system. We made a prototype system with two 120Hz displays and successfully showed four viewpoints.

A super multi-view head-up display (SMV-HUD) was developed. The smooth motion parallax provided by the SMV technique enables a precise superposition of three-dimensional (3D) images on real objects. The developed SMV-HUD was used to explore display conditions to provide smooth motion parallax. It had three configurations that display 3D images in short-, medium-, and long-distance ranges, assuming the supposed usage of PC monitors, TVs, and public viewing, respectively. The subjective evaluation was performed by changing the depth of 3D images and the interval of viewing points. The interval of viewing points was changed by displaying identical parallax images to succeeding viewing points. We found that the ratio of the image shift between adjacent parallax images to the pixel pitch of 3D images dominated the perception of the unnatural motion parallax. When the ratio was smaller than 0.2, the discontinuity was not perceived. When the ratio was larger than 1, the discontinuity was always perceived and the 3D resolution decreased two times at transition points of viewing points. When the crosstalk between viewing points was relatively large, the discontinuity was not perceived even when the ratio was one or two, although the resolution decreased two or three times throughout the viewing region.

As the viewing safety issues in stereoscopic 3D services have been under the spotlight again, it has been more important to investigate determinants of visual discomfort in viewing stereoscopic images. In general, excessive binocular disparity has been regarded as one of key determinants of visual discomfort in stereoscopic viewing. However, in consideration of the complexity of the visual system, the degree of perceived visual discomfort could be also different depending on other characteristics of visual stimulus. Inspired by previous studies that have investigated the relation between stimulus width and binocular fusion limit, we assume that stimulus width can also affect the subjective sensation of visual discomfort in stereoscopic viewing. This paper investigates the relationship between stimulus width and visual comfort by measuring subjective visual discomfort. Experimental results showed that smaller stimulus width could induce more visual discomfort.

We evaluated an observer’s fatigue and a sense of presence, in observing the no-disparity realistic image (NDR image) [1-3]. NDR image is consisted of two images (right and left image). Right image is created by shifting all pixels in left image same amount. Consequently, there are no disparities in all area of NDR image. NDR image which is reconfigured the contents that contain image with large disparity might have a possibility that it suppress an observer’s fatigue and let him feel high presence. Subjects observed three condition's videos (stereoscopic, NDR and 2D). Subjects observed two videos in each condition. Each video was 30 minutes. There were scenes with large disparity (more than 5 degrees) in videos. Subjects responded SSQ (Simulator Sickness Questionnaire), VAS (Visual Analogue Scale) for fatigue and questionnaire on realism and were measured CFF (Critical Frequency Fusion), accommodation tremor, stereoscopic vision test, ocular position measurement and eye sight test. Results showed that NDR image let observer feel high presence and an observer's fatigue was low. NDR image is effective, even if contents which contain large disparity are converted into NDR image.

The pair comparison method is often recommended in subjective experiments because of the reliability of the obtained results. However, a drawback of this method is that the number of comparisons increases exponentially with the number of stimuli, which limits its usability for a large number of stimuli. Several design methods that aim to reduce the number of comparisons were proposed in the literature. However, their performances in the context of 3DTV should be evaluated carefully due to the fact that the results obtained from a paired comparison experiment in 3DTV may be influenced by two important factors. One is the observation error from observer's attentiveness, in particular inverting the vote. The second factor concerns the dependence on the context in which the evaluation takes place. In this study, three design methods, namely Full Paired Comparison method (FPC), Square Design method (SD) and the Adaptive Square Design method (ASD) were evaluated by subjective visual discomfort experiment in 3DTV. The results from the FPC method were considered as the ground truth. Comparing with the ground truth, the ASD method provided the most accurate results with a given number of trials. It also showed the highest robustness against observation errors and interdependence of comparisons. Due to the efficiency of the ASD method, paired comparison experiments become feasible with a reasonably large number of stimuli for measuring 3DTV visual discomfort.

There are different panoramic techniques to produce outstanding stereoscopic panoramas of static scenes. However, a camera configuration capable to capture omnidirectional stereoscopic snapshots and videos of dynamic scenes is still a subject of research. In this paper, two multiple-camera configurations capable to produce high-quality stereoscopic panoramas in real-time are presented. Unlike existing methods, the proposed multiple-camera systems acquire all the information necessary to render stereoscopic panoramas at once. The first configuration exploits micro-stereopsis arising from a narrow baseline to produce omni-stereoscopic images. The second panoramic camera uses an extended baseline to produce poly-centric panoramas and to extract additional depth information, e.g., disparity and occlusion maps, which are used to synthesize stereoscopic views in arbitrary viewing directions. The results of emulating both cameras and the pros and cons of each set-up are presented in this paper.

There are three main approaches creating stereoscopic S3D content: stereo filming using two cameras, stereo rendering of 3D computer graphics, and 2D to S3D conversion by adding binocular information to 2D material images. Although manual “off-line” conversion can control the amount of parallax flexibly, 2D material images are converted according to monocular information in most cases, and the flexibility of 2D to S3D conversion has not been exploited. If the depth is expressed flexibly, comprehensions and interests from converted S3D contents are anticipated to be differed from those from 2D. Therefore, in this study we created new S3D content for education by applying 2D to S3D conversion. For surgical education, we created S3D surgical operation content under a surgeon using a partial 2D to S3D conversion technique which was expected to concentrate viewers’ attention on significant areas. And for art education, we converted Ukiyoe prints; traditional Japanese artworks made from a woodcut. The conversion of this content, which has little depth information, into S3D, is expected to produce different cognitive processes from those evoked by 2D content, e.g., the excitation of interest, and the understanding of spatial information. In addition, the effects of the representation of these contents were investigated.

In this study, for the heritage of industrial modernization we produced a stereoscopic archive of the JVC KENWOOD Yokohama Factory façade, which was dismantled in March 2011, by converting 2D into 3D images. Further, drawn images of the cultural asset were also converted into 3D images in order to evaluate them as part of a stereoscopic archive in terms of how well they express the asset. An experiment was conducted to compare subjects' content impressions under the different image conditions. For the experiment, head mounted display and 42-inch 3D TV were used. 30 students in twenties years of age. with normal binocular vision. participated through questionnaires and interviews to compare the impressions, between the conditions.

In this paper we present a novel image quality assessment technique for evaluating virtual synthesized views in the context of multi-view video. In particular, Free Viewpoint Videos are generated from uncompressed color views and their compressed associated depth maps by means of the View Synthesis Reference Software, provided by MPEG. Prior to the synthesis step, the original depth maps are encoded with different coding algorithms thus leading to the creation of additional artifacts in the synthesized views. The core of proposed wavelet-based metric is in the registration procedure performed to align the synthesized view and the original one, and in the skin detection that has been applied considering that the same distortion is more annoying if visible on human subjects rather than on other parts of the scene. The effectiveness of the metric is evaluated by analyzing the correlation of the scores obtained with the proposed metric with Mean Opinion Scores collected by means of subjective tests. The achieved results are also compared against those of well known objective quality metrics. The experimental results confirm the effectiveness of the proposed metric.

There are several studies on estimating crosstalk of 3D displays. Crosstalk is an important factor in determining image quality of stereoscopic display. In previous study, gray to gray crosstalk model has been modified. In this paper, we use six commercial stereoscopic displays including passive polarized glasses and active shutter glasses, both left and right eyes were measured by a luminance meter to verify the gray to gray crosstalk model and analyze the stability of the measurement data using our standard operating procedure (SOP). According to the SOP, we use simple statistical method to verify the repeatability of data. Experimental results indicate that our measurement system can be used to estimate the value of gray to gray crosstalk of stereoscopic display, the uncertainty below 0.8% for FPR display and 3% for shutter glasses display at 95% confidence level. These results, can be used as an important parameters in stereoscopic display visual quality, also can be used as a design feedback for engineer.

Frequency features of stereo images are investigated in the DFT (Discrete Fourier Transform) domain by characterizing phase and magnitude properties originated from the horizontal disparities in the stereo images. Also, the well-known DFT properties including the conjugate symmetry property are utilized to identify essential frequency components of stereo images. Our investigation reveals that the DFT of the stereo images has useful properties that can prioritize the DFT coefficients for compact representations and compressions.

A depth dilation filter is proposed for free viewpoint video system based on mixed resolution multi-view video plus depth (MVD). By applying gray scale dilation filter to depth images, foreground regions are extended to background region, and synthesis artifacts occur out of boundary edge. Thus, objective and subjective quality of view synthesis result is improved. A depth dilation filter is applied to inloop resampling part in encoding/decoding, and post processing part after decoding. Accurate view synthesis is important in virtual view generation for autostereoscopic display, moreover there are many coding tools which use view synthesis to reduce interview redundancy in 3D video coding such as view synthesis prediction (VSP) and depth based motion vector prediction (DMVP), and compression efficiency can be improved by accurate view synthesis. Coding and synthesis experiments are performed for performance evaluation of a dilation filter with MPEG test sequences. Dilation filter was implemented on the top of the MPEG reference software for AVC based 3D video coding. By applying a depth dilation filter, BD-rate gains of 0.5% and 6.0% in terms of PSNR of decoded views and synthesized views, respectively.

In this paper, we discuss a Free viewpoint synthesis with View + Depth format for Multiview applications such as 3DTV and Free View-point Television(FTV)¹.When generating a virtual image, 3D warping is used with view and depth of a reference camera. This process includes the problem that holes appear in the virtual image. In conventional method, the holes were dealt with collectively by median filter. There are some different reasons why holes appear through this process. So, it is improper that they are not distinguished particularly and treated all at once like conventional method. We analyze the factors, and recognize that two ones exist, boundary between foreground and background, and reduction of resolution. In this paper, we propose a new hole filling method considering these factors. In the first step, we classify nearby pixels into boundary or same object area according to the gradient of depth value. For boundary case, we hold them and refer to other two real cameras. For another case of same object area, we set up sub-pixels between nearby pixels and warp them if the depth is gradually changing or virtual viewpoint of the warped image is closer to the object than the original view position because they probably cause holes from reduction of resolution. We implement these methods in the simulation. As a result, we prevent boundary in the virtual image from being ambiguous, and confirm the availability of proposed method.