Details Of Scene Research Articles

3D photon counting integral imaging with unknown sensor positions

Photon counting techniques have been introduced with integral imaging for three-dimensional (3D) imaging applications. The previous reports in this area assumed a priori knowledge of exact sensor positions for 3D image reconstruction, which may be difficult to satisfy in certain applications. In this paper, we extend the photon counting 3D imaging system to situations where sensor positions are unknown. To estimate sensor positions in photon counting integral imaging, scene details of photon counting images are needed for image correspondences matching. Therefore, an iterative method based on the total variation maximum a posteriori expectation maximization (MAP-EM) algorithm is used to restore photon counting images. Experimental results are presented to show the feasibility of the method. To the best of our knowledge, this is the first report on 3D photon counting integral imaging with unknown sensor positions.

Experimental forensic scenes for the characterization of ground-penetrating radar wave response

Over the last few decades, the use of non-intrusive geophysical techniques, which allows for coverage of an entire crime scene in a reasonable amount of time, in forensic investigation has increased. In this study, we analyze the effectiveness of ground-penetrating radar (GPR) in forensics. Experimental scenes were simulated and some of the most commonly buried items in actual crime scenes were introduced, such as bone remains, guns and drug caches. Later, a GPR survey was conducted on the experimental grids with a 500 MHz antenna. The final purpose was to characterize the radar wave response expected for each set of remains to assist with its identification in later actual investigations. The results collected provided promising information that can be used when surveying real cases. Nevertheless, there were some interpretational difficulties regarding the sizes of the items and the electromagnetic properties of the materials. For these cases, finite-difference time-domain modeling was employed to achieve an advanced interpretation of the field data. The simulated models used were built from accurate geometric data provided by photogrammetric methods, which replicate the experimental scenes in fine detail. Furthermore, this approach allowed for the simulation of more realistic models, and the synthetic data obtained provided valuable information for assisting in the interpretation of field data. As a result of this work, it was concluded that GPR can be an effective tool when searching for a variety of materials during a crime scene investigation.

Adaptive Image Registration via Hierarchical Voronoi Subdivision

Advances in image acquisition systems have made it possible to capture high-resolution images of a scene, recording considerable scene details. With increased resolution comes increased image size and geometric difference between multiview images, complicating image registration. Through Voronoi subdivision, we subdivide large images into small corresponding regions, and by registering small regions, we register the images in a piecewise manner. Image subdivision reduces the geometric difference between regions that are registered and simplifies the correspondence process. The proposed method is a hierarchical one. While previous methods use the same block size and shape at a hierarchy, the proposed method adapts the block size and shape to the local image details and geometric difference between the images. This adaptation makes it possible to keep geometric difference between corresponding regions small and simplifies the correspondence process. Implementational details of the proposed image registration method are provided, and experimental results on various types of images are presented and analyzed.

A 3D Perspective Display Cube for Multiview Tracking

This paper describes an eye naked 3D display cube, which can track a viewer’s position and render the stereo perspective scene in each display surface. The viewers can look the details of 3D models and scenes from any viewpoints. It seems as the real 3D objects lay in the display cube. It supports multi-viewer interactive in one scene. The viewers also can see the position of each other in own screen. They can co-operate in the same scene, although their field of view is independent of each other. The technologies of 3D perspective rendering, marker tracking and image processing are used for the 3D display cube.

Utilizing Clinical Eye Models to Predict Retinal Image Quality of Individuals

A method of diminishing the shape error of membrane mirror is proposed in this paper. The inner inflating pressure is considerably decreased by adopting the pre-shaped membrane. Small deformation of the membrane mirror with greatly reduced shape error is sequentially achieved. Primarily a finite element model of the above pre-shaped membrane is built on the basis of its mechanical properties. Then accurate shape data under different pressures can be acquired by iteratively calculating the node displacements of the model. Shape data are applicable to build up deformed reflecting surfaces for the simulative analysis in ZEMAX. Finally, ground-based imaging experiments of 4-bar targets and nature scene are conducted. Experiment results indicate that the MTF of the infrared system can reach to 0.3 at a high spatial resolution of 10l p/mm, and texture details of the nature scene are well-presented. The method can provide theoretical basis and technical support for the applications in lightweight optical components with ultra-large apertures.

Probabilistic Exposure Fusion

The luminance of a natural scene is often of high dynamic range (HDR). In this paper, we propose a new scheme to handle HDR scenes by integrating locally adaptive scene detail capture and suppressing gradient reversals introduced by the local adaptation. The proposed scheme is novel for capturing an HDR scene by using a standard dynamic range (SDR) device and synthesizing an image suitable for SDR displays. In particular, we use an SDR capture device to record scene details (i.e., the visible contrasts and the scene gradients) in a series of SDR images with different exposure levels. Each SDR image responds to a fraction of the HDR and partially records scene details. With the captured SDR image series, we first calculate the image luminance levels, which maximize the visible contrasts, and then the scene gradients embedded in these images. Next, we synthesize an SDR image by using a probabilistic model that preserves the calculated image luminance levels and suppresses reversals in the image luminance gradients. The synthesized SDR image contains much more scene details than any of the captured SDR image. Moreover, the proposed scheme also functions as the tone mapping of an HDR image to the SDR image, and it is superior to both global and local tone mapping operators. This is because global operators fail to preserve visual details when the contrast ratio of a scene is large, whereas local operators often produce halos in the synthesized SDR image. The proposed scheme does not require any human interaction or parameter tuning for different scenes. Subjective evaluations have shown that it is preferred over a number of existing approaches.

The Trouble with History and Fiction

The Trouble with History and Fiction

Generalized Random Walks for Fusion of Multi-Exposure Images

A single captured image of a real-world scene is usually insufficient to reveal all the details due to under- or over-exposed regions. To solve this problem, images of the same scene can be first captured under different exposure settings and then combined into a single image using image fusion techniques. In this paper, we propose a novel probabilistic model-based fusion technique for multi-exposure images. Unlike previous multi-exposure fusion methods, our method aims to achieve an optimal balance between two quality measures, i.e., local contrast and color consistency, while combining the scene details revealed under different exposures. A generalized random walks framework is proposed to calculate a globally optimal solution subject to the two quality measures by formulating the fusion problem as probability estimation. Experiments demonstrate that our algorithm generates high-quality images at low computational cost. Comparisons with a number of other techniques show that our method generates better results in most cases.

The “Parahippocampal Place Area” Responds Preferentially to High Spatial Frequencies in Humans and Monkeys

Author SummaryMany reports suggest that different categories of visual stimuli are processed in correspondingly specific “modules” in the visual cortex. For instance, images of faces are processed in one cortical module (the “fusiform face area”), while images of scenes are processed in an adjacent module (the “parahippocampal place area,” or PPA). How does the PPA encode for such high-level, complex visual scenes? In this study, we show that at least part of the PPA response is due to a lower-level variable, reflected as higher spatial frequencies. These are prominent in the edges and details of scenes, but less prominent in faces and other stimuli. When we altered standard images of faces and places so that they only contained low, medium, or high spatial frequencies, we found that the PPA responded strongly to images containing high spatial frequencies. Importantly, using the same stimuli as for the human studies, we also demonstrated a homolog of human PPA in macaque temporal cortex (“mPPA”). As in humans, mPPA responds selectively to higher spatial frequencies. This demonstration of PPA in macaques paves the way for carrying out further electrophysiological and anatomical studies that may help elucidate the neural mechanisms for place selectivity in the human visual cortex.

Attention-driven monocular scene reconstruction for obstacle detection, robot navigation and map building

In this paper, we present a feature-based approach for monocular scene reconstruction based on Extended Kalman Filters (EKF). Our method processes a sequence of images taken by a single camera mounted frontally on a mobile robot. Using a combination of various techniques, we are able to produce a precise reconstruction that is free from outliers and can therefore be used for reliable obstacle detection and 3D map building. Furthermore, we present an attention-driven method that focuses the feature selection to image areas where the obstacle situation is unclear and where a more detailed scene reconstruction is necessary. In extensive real-world field tests we show that the presented approach is able to detect obstacles that are not seen by other sensors, such as laser range finders. Furthermore, we show that visual obstacle detection combined with a laser range finder can increase the detection rate of obstacles considerably, allowing the autonomous use of mobile robots in complex public and home environments.

Emotion enhances the subjective feeling of remembering, despite lower accuracy for contextual details.

Emotion strengthens the subjective experience of recollection. However, these vivid and confidently remembered emotional memories may not necessarily be more accurate. We investigated whether the subjective sense of recollection for negative stimuli is coupled with enhanced memory accuracy for contextual details using the remember/know paradigm. Our results indicate a double-dissociation between the subjective feeling of remembering, and the objective memory accuracy for details of negative and neutral scenes. "Remember" judgments were boosted for negative relative to neutral scenes. In contrast, memory for contextual details and associative binding was worse for negative compared to neutral scenes given a "remember" response. These findings show that the enhanced subjective recollective experience for negative stimuli does not reliably indicate greater objective recollection, at least of the details tested, and thus may be driven by a different mechanism than the subjective recollective experience for neutral stimuli.

The role of semantic interference in limiting memory for the details of visual scenes

Many studies suggest a large capacity memory for briefly presented pictures of whole scenes. At the same time, visual working memory (WM) of scene elements is limited to only a few items. We examined the role of retroactive interference in limiting memory for visual details. Participants viewed a scene for 5 s and then, after a short delay containing either a blank screen or 10 distracter scenes, answered questions about the location, color, and identity of objects in the scene. We found that the influence of the distracters depended on whether they were from a similar semantic domain, such as “kitchen” or “airport.” Increasing the number of similar scenes reduced, and eventually eliminated, memory for scene details. Although scene memory was firmly established over the initial study period, this memory was fragile and susceptible to interference. This may help to explain the discrepancy in the literature between studies showing limited visual WM and those showing a large capacity memory for scenes.

Prism adaptation reverses the local processing bias in patients with right temporo-parietal junction lesions

Lesions to the right temporo-parietal cortex commonly result in hemispatial neglect. Lesions to the same area are also associated with hyperattention to local details of a scene and difficulty perceiving the global structure. This local processing bias is an important factor contributing to neglect and may contribute to the higher prevalence of the disorder following right compared with left hemisphere strokes. In recent years, visuomotor adaptation to rightward-shifting prisms has been introduced as a promising treatment for hemispatial neglect. Explanations for these improvements have generally described a leftward realignment of attention, however, the present investigation provides evidence that prism adaptation reduces the local processing bias. Five patients with right temporal-parietal junction lesions were asked to identify the global or local levels of hierarchical figures before and after visuomotor adaptation to rightward-shifting prisms. Prior to prism adaptation the patients had difficulty ignoring the local elements when identifying the global component. Following prism adaptation, however, this pattern was reversed, with greater global interference during local level identification. The results suggest that prism adaptation may improve non-spatially lateralized deficits that contribute to the neglect syndrome.

A Biologically Inspired Smart Camera for Use in Surveillance Applications

Biological vision systems are capable of discerning detail as well as detecting objects and motion in a wide range of highly variable lighting conditions that proves challenging to traditional cameras. In this paper, the authors describe the real-time implementation of a biological vision model using a high dynamic range video camera and a General Purpose Graphics Processing Unit. The effectiveness of this implementation is demonstrated in two surveillance applications: dynamic equalization of contrast for improved recognition of scene detail and the use of biologically-inspired motion processing for the detection of small or distant moving objects in a complex scene. A system based on this prototype could improve surveillance capability in any number of difficult situations.

Estimating perception of scene layout properties from global image features

The relationship between image features and scene structure is central to the study of human visual perception and computer vision, but many of the specifics of real-world layout perception remain unknown. We do not know which image features are relevant to perceiving layout properties, or whether those features provide the same information for every type of image. Furthermore, we do not know the spatial resolutions required for perceiving different properties. This paper describes an experiment and a computational model that provides new insights on these issues. Humans perceive the global spatial layout properties such as dominant depth, openness, and perspective, from a single image. This work describes an algorithm that reliably predicts human layout judgments. This model's predictions are general, not specific to the observers it trained on. Analysis reveals that the optimal spatial resolutions for determining layout vary with the content of the space and the property being estimated. Openness is best estimated at high resolution, depth is best estimated at medium resolution, and perspective is best estimated at low resolution. Given the reliability and simplicity of estimating the global layout of real-world environments, this model could help resolve perceptual ambiguities encountered by more detailed scene reconstruction schemas.

Accumulating and Remembering the Details of Neutral and Emotional Natural Scenes

In contrast to our rich sensory experience with complex scenes in everyday life, the capacity of visual working memory is thought to be quite limited. Here our memory has been examined for the details of naturalistic scenes as a function of display duration, emotional valence of the scene, and delay before test. Individual differences in working memory and long-term memory for pictorial scenes were examined in experiment 1. The accumulation of memory for emotional scenes and the retention of these details in long-term memory were investigated in experiment 2. Although there were large individual differences in performance, memory for scene details generally exceeded the traditional working memory limit within a few seconds. Information about positive scenes was learned most quickly, while negative scenes showed the worst memory for details. The overall pattern of results was consistent with the idea that both short-term and long-term representations are mixed together in a medium-term 'online' memory for scenes.

Extended use of incremental signal-to-noise ratio as reliability criterion for multiple-slope wide-dynamic-range image capture

Mobile applications present new image quality challenges. Automotive vision requires reliable capture of scene detail. Photospace measurements have shown that the extremely wide intrascene dynamic range of traffic scenes necessitates wide-dynamic-range (WDR) technology. Multiple-slope complementary metal-oxide semiconductor (CMOS) technology adaptively extends dynamic range by partially resetting the pixel, resulting in a response curve with piecewise linear slopes of progressively increasing compression. As compression and thus dynamic range increase, a trade-off against detail loss is observed. Incremental signal-to-noise ratio (iSNR) has been proposed in ISO/TC42 standards for determining dynamic range, and this work describes how to adapt these to WDR. Measurements and computer simulations reveal that the observed trade-off between WDR extension and the loss of local detail can be explained by a drop in iSNR at each reset point. If a reset is not timed optimally, then iSNR may drop below the detection limit causing an iSNR hole to appear within the dynamic range. Thus iSNR has extended utility: it not only determines the dynamic range limits but also defines dynamic range as the luminance range where detail detection is reliable. It has become the critical criterion when maximizing dynamic range to maintain the minimum necessary level of detection reliability.

Observations on increased accidental asphyxia deaths in infancy while cosleeping in the state of Maryland.

The Office of the Chief Medical Examiner (OCME) has recorded a significant increase of accidental asphyxia deaths in infancy associated with cosleeping in the state of Maryland in 2003. A total of 102 infants died suddenly and unexpectedly during 2003 in the state of Maryland. Of the 102 infants, 46 (45%) were found cosleeping. The frequency of cosleeping among these 102 infants was 28% (29/102) for black infants and 15% (15/102) for white infants. Ten of the 46 cosleeping infant deaths (20%) were determined to be the result of accidental asphyxia, and 28 cosleeping infant deaths (59%) were classified as "undetermined" because the possibility of asphyxia due to overlay while cosleeping could not be ruled out. Only 21 cases were determined to be Sudden Infant Death Syndrome (SIDS), which is consistent with the continuous decline of SIDS death in Maryland since 1994. The age of asphyxiated cosleeping infants ranged from 15 days to 9 months. Nine out of the 10 asphyxia deaths were black infants. The most common sleeping location of the asphyxia infants was on a couch/sofa, followed by an adult bed. Crib availability was documented in all of the cosleeping cases. A majority (61%) of the cosleeping infants (28/46) had an available crib or bassinet at home and 9 out of 10 asphyxiated cosleeping infants had a crib at home at the time of the incident. This report focuses on the detailed scene investigation findings of infant victims who died of asphyxia while cosleeping. The shift of diagnosis in sudden infant death investigation is also addressed.

Gradient‐Preserving Color Transfer

AbstractColor transfer is an image processing technique which can produce a new image combining one source image's contents with another image's color style. While being able to produce convincing results, however, Reinhard et al.'s pioneering work has two problems—mixing up of colors in different regions and the fidelity problem. Many local color transfer algorithms have been proposed to resolve the first problem, but the second problem was paid few attentions.In this paper, a novel color transfer algorithm is presented to resolve the fidelity problem of color transfer in terms of scene details and colors. It's well known that human visual system is more sensitive to local intensity differences than to intensity itself. We thus consider that preserving the color gradient is necessary for scene fidelity. We formulate the color transfer problem as an optimization problem and solve it in two steps—histogram matching and a gradient‐preserving optimization. Following the idea of the fidelity in terms of color and gradient, we also propose a metric for objectively evaluating the performance of example‐based color transfer algorithms. The experimental results show the validity and high fidelity of our algorithm and that it can be used to deal with local color transfer.

Eye‐movement patterns are associated with communicative competence in autistic spectrum disorders

Investigations using eye-tracking have reported reduced fixations to salient social cues such as eyes when participants with autism spectrum disorders (ASD) view social scenes. However, these studies have not distinguished different cognitive phenotypes. The eye-movements of 28 teenagers with ASD and 18 typically developing peers were recorded as they watched videos of peers interacting in familiar situations. Within ASD, we contrasted the viewing patterns of those with and without language impairments. The proportion of time spent viewing eyes, mouths and other scene details was calculated, as was latency of first fixation to eyes. Finally, the association between viewing patterns and social-communicative competence was measured. Individuals with ASD and age-appropriate language abilities spent significantly less time viewing eyes and were slower to fixate the eyes than typically developing peers. In contrast, there were no differences in viewing patterns between those with language impairments and typically developing peers. Eye-movement patterns were not associated with social outcomes for either language phenotype. However, increased fixations to the mouth were associated with greater communicative competence across the autistic spectrum. Attention to both eyes and mouths is important for language development and communicative competence. Differences in fixation time to eyes may not be sufficient to disrupt social competence in daily interactions. A multiple cognitive deficit model of ASD, incorporating different language phenotypes, is advocated.