Disparity Computation Research Articles

Monocular transparency and unpaired stereopsis

Howard and Duke [Howard, I. P. & Duke, P. A. (2003). Monocular transparency generates quantitative depth. Vision Research, 43, 2615–2621] recently proposed a new source of binocular information they claim is used to recover depth in stereoscopic displays. They argued that these displays lack conventional disparity and that the metrical depth experienced results from transparency rather than occlusion relations. Using a variety of modified versions of their stimuli, we show here that the conditions for transparency are not required to elicit the depth experienced in their stereograms. We demonstrate that quantitative and precise depth depended not on the presence of transparency but horizontal contours of the same contrast polarity. Depth was attenuated, particularly at larger target offsets, when horizontal contours had opposite contrast polarity for at least a portion of their length. We also show that a demonstration they used to control for the role of horizontal contours can be understood with previously identified mechanisms involved in the computations associated with stereoscopic occlusion. These results imply that the findings reported by Howard and Duke can be understood with mechanisms responsible for the computation of binocular disparity and stereoscopic occlusion.

A Data Management Approach to Quality Assurance Quality Assurance Using Colorectal Carcinoma Reports From Two Institutionsas a Model

We used data management software to compare pathology report data concerning regional lymph node sampling for colorectal carcinoma from 2 institutions using different dissection methods. Data were retrieved from 2 disparate anatomic pathology information systems for all cases of colorectal carcinoma in 2003 involving the ascending and descending colon. Initial sorting of the data included overall lymph node recovery to assess differences between the dissection methods at the 2 institutions. Additional segregation of the data was used to challenge the application's capability of accurately addressing the complexity of the process. This software approach can be used to evaluate data from disparate computer systems, and we demonstrate how an automated function can enable institutions to compare internal pathologic assessment processes and the results of those comparisons. The use of this process has future implications for pathology quality assurance in other areas.

What and where in the computation of relative disparity

What and where in the computation of relative disparity

A coarse-to-fine disparity energy model with both phase-shift and position-shift receptive field mechanisms.

Numerous studies suggest that the visual system uses both phase- and position-shift receptive field (RF) mechanisms for the processing of binocular disparity. Although the difference between these two mechanisms has been analyzed before, previous work mainly focused on disparity tuning curves instead of population responses. However, tuning curve and population response can exhibit different characteristics, and it is the latter that determines disparity estimation. Here we demonstrate, in the framework of the disparity energy model, that for relatively small disparities, the population response generated by the phase-shift mechanism is more reliable than that generated by the position-shift mechanism. This is true over a wide range of parameters, including the RF orientation. Since the phase model has its own drawbacks of underestimating large stimulus disparity and covering only a restricted range of disparity at a given scale, we propose a coarse-to-fine algorithm for disparity computation with a hybrid of phase-shift and position-shift components. In this algorithm, disparity at each scale is always estimated by the phase-shift mechanism to take advantage of its higher reliability. Since the phase-based estimation is most accurate at the smallest scale when the disparity is correspondingly small, the algorithm iteratively reduces the input disparity from coarse to fine scales by introducing a constant position-shift component to all cells for a given location in order to offset the stimulus disparity at that location. The model also incorporates orientation pooling and spatial pooling to further enhance reliability. We have tested the algorithm on both synthetic and natural stereo images and found that it often performs better than a simple scale-averaging procedure.

Dissecting application security XML schemas

Recent activities in applications security research and industry have inspired companies and researchers to create a standard way of defining and describing vulnerabilities, expressing different application needs with respect to various aspects of vulnerabilities, making vulnerability descriptions accessible to disparate computer systems. This article describes the state of the art today and possible directions in the future. It summarizes industry efforts and focuses on application security related XML schemas being developed within OASIS.

Single-Chip Stereo Imager

A stereo vision chip, incorporating two 128 /spl times/ 128 pixel current-mode imagers and analog disparity computation circuitry, is presented. A modified version of block matching is used to compute the disparity between the two images at each (u, v) coordinate, with the sum-of-absolute-difference value for each possible disparity being computed in parallel. The chip has been tested at computation rates up to 11.2 million checked disparities per second, while consuming only 35 mW from a 5 V supply, including imagers and computation circuits.

Stereoscopic depth perception from oblique phase disparities

In order to understand the role of oblique retinal image disparities in the perception of stereoscopic depth, we measured the depth perceived from random dot stereograms in which phase disparities were introduced in a selected band of stimulus orientations. A band of orientation was defined by a center orientation that ranged from 7.5 (near vertical) to 82.5 o[rientation]deg and by a bandwidth that was defined as the difference between the highest and the lowest orientation in the band. The bandwidths tested were 15, 30 and 45 odeg. A constant phase disparity of 90 p[hase]deg was introduced in all of the oriented spatial frequency components within the orientation band and the perceived depth of each stimulus was matched using a small square binocular probe. For each bandwidth, perceived depth increased with an increase in the center orientation up to approximately 60 odeg. This suggests that the human stereovision system derives a large proportion of information about perceived stereoscopic depth from oblique phase disparities. Simulations using an energy model of stereoscopic depth perception indicate that oblique phase disparities are unlikely to be processed by neural mechanisms tuned to near-vertical orientations within the stimulus. Our results therefore suggest that oblique retinal disparities are initially detected as oblique phase disparities by binocular mechanisms tuned to oblique orientations. Because the perceived depth from oblique phase disparities is consistent with the trigonometrically determined equivalent horizontal disparities, we presume that the information from oblique phase disparities is included in the visual system’s computation of the horizontal retinal disparity.

Reading a population code: a multi-scale neural model for representing binocular disparity

Although binocular neurons in the primary visual cortex are sensitive to retinal disparity, their activity does not constitute an unambiguous disparity signal. A multi-spatial-scale neural model for disparity computation is developed to examine how population activity might be interpreted to overcome ambiguities at the single neuron level. The model incorporates a front end that encodes disparity by a family of complex cell-like energy units and a second stage that reads the population activity. Disparity is recovered by matching the population response to a set of canonical templates, derived from the mean response to white noise stimuli at a range of disparities. Model predictions are qualitatively consistent with a variety of psychophysical results in the literature, including the effects of spatial frequency on stereoacuity and bias in perceived depths, and the effect of standing disparity on increment thresholds. Model predictions are also consistent with data on qualitative appearance of complex stimuli, including depth averaging, transparency, and corrugation. The model also accounts for the non-linear interaction of disparities in compound grating stimuli. These results show that a template-match approach reduces ambiguities in individual and pooled neuronal responses, and allows for a broader range of percepts, consistent with psychophysics, than other models. Thus, the pattern of neural population activity across spatial scales is a better candidate for the neural correlate of depth perception than the activity of single neurons or the pooled activity of multiple neurons.

A self-describing data transfer model for ITS applications

The wide variety of remote sensors used in Intelligent Transportation Systems (ITS) applications (loops, probe vehicles, radar, cameras, etc.) has created a need for general methods by which data can be shared among agencies and users who own disparate computer systems. In this paper, we present a methodology that demonstrates that it is possible to create, encode, and decode a self-describing data stream using: 1) existing data description language standards; 2) parsers to enforce language compliance; 3) a simple content language that flows out of the data description language; and 4) architecture neutral encoders and decoders based on ASN.1.

A simple mechanism for sharing and transporting medical digital case information across disparate computer language and data storage environments.

It is challenging to remotely share generic medical case information without an agreed upon definition of a medical digital teaching file (DTF). By utilizing an application of the extensible markup language (XML) called web-distributed data exchange (WDDX) along with an agreed upon WDDX structure, it is technically easy to share or syndicate medical case DTFs across computing environments that use different information models and computer languages. Thus, this easily implemented technology offers us an immediately available means to share and increase the value of scientific knowledge.

Motion direction, speed and orientation in binocular matching.

The spatial differences between the images seen by the two eyes, called binocular disparities, can be used to recover the volumetric (three-dimensional) aspects of a scene. The computation of disparity depends upon the correct identification of corresponding features in the two images. Understanding what image features are used by the brain to solve this matching problem is one of the main issues in stereoscopic vision. Many cortical neurons in visual areas V1 (ref. 2), MT (refs 3, 4) and MST (refs 5, 6) that are tuned to binocular disparity are also tuned to orientation, motion direction and speed. Although psychophysical work has shown that motion direction can facilitate binocular matching, the psychophysical literature on the role of orientation is mixed, and it has been argued that speed differences are ineffective in aiding correspondence. Here we use a different psychophysical paradigm to show that the visual system uses similarities in orientation, motion direction and speed to achieve binocular correspondence. These results indicate that cells that multiplex orientation, motion direction, speed and binocular disparity may help to solve the binocular matching problem.

Making the Cisco connection: the story behind the real Internet superpower

Making the Cisco connection: the story behind the real Internet superpower

N Degrees of Separation: Influences of Dynamic Programming on Computer Science

Bellman's dynamic programming methodology can be applied to a wide range of computer science optimization problems. Some of these applications are briefly reviewed here. This work has led to advances in numerous other areas of computer science, including programming languages, computer simulation, and computer architecture. In turn, advances in these other areas suggest other applications of and to dynamic programming. In summary, we survey interrelationships of disparate computer science areas, and how progress in these areas evolved from Bellman's work.

A physiologically-based explanation of disparity attraction and repulsion

Westheimer and Levi [(1987) Vision Research, 27, 1361–1368] found that when a few isolated features are viewed foveally, the perceived depth of a feature depends not only on its own disparity but also on those of its neighbors. The nature of this interaction is a function of the lateral separation between the features: When the distance is small the features appear to attract each other in depth but the interaction becomes repulsive at larger distances. Here we introduce a two-dimensional extension of our recent stereo model based on the physiological studies of Ohzawa, DeAngelis and Freeman [(1990) Science, 249, 1037–1041] and demonstrate through analyses and simulations that these observations can be naturally explained without introducing ad hoc assumptions about the connectivity between disparity-tuned units. In particular, our model can explain the distance-dependent attraction/repulsion phenomena in both the vertical-line configuration used by Westheimer [(1986) Journal for Neurophysiology, 370, 619–629], and the horizontal-line-and-point configuration used by Westheimer and Levi. Thus, the psychophysically observed disparity interaction may be viewed as a direct consequence of the known physiological organization of the binocular receptive fields. We also find that the transition distance at which the disparity interaction between features changes from attraction to repulsion is largely determined by the preferred spatial frequency and orientation distributions of the cells used in the disparity computation. This result may explain the observed variations of the transition distance among different subjects in the psychophysical experiments. Finally, our model can also reproduce the observed effect on the perceived disparity when the disparity magnitude of the neighboring features is changed.

Relationship between phase and energy methods for disparity computation.

The phase and energy methods for computing binocular disparity maps from stereograms are motivated differently, have different physiological relevances, and involve different computational steps. Nevertheless, we demonstrate that at the final stages where disparity values are made explicit, the simplest versions of the two methods are exactly equivalent. The equivalence also holds when the quadrature-pair construction in the energy method is replaced with a more physiologically plausible phase-averaging step. The equivalence fails, however, when the phase-difference receptive field model is replaced by the position-shift model. Additionally, intermediate results from the two methods are always quite distinct. In particular, the energy method generates a distributed disparity representation similar to that found in the visual cortex, while the phase method does not. Finally, more elaborate versions of the two methods are in general not equivalent. We also briefly compare these two methods with some other stereo models in the literature.

A model for intradendritic computation of binocular disparity.

Many complex cells in mammalian primary visual cortex are finely tuned to binocular disparity. In the prevailing model, several binocular simple cells drive each disparity-tuned complex cell. However, some cat complex cells receive direct LGN input, and binocular simple cells are rare in macaque. In our biophysically detailed compartmental model, active dendrites of a single neuron perform the multiple simple-cell-like subunit computations that underlie both orientation and disparity tuning. The responses of our detailed model could be predicted by a simple algebraic formula closely related to an 'energy' model. Adding inhibitory synapses led to sharper, more contrast-invariant tuning curves. Thus active dendrites could contribute to binocular-disparity tuning in complex cells.

A hybrid approach for stereo disparity computation

An alternative, hybrid approach for disparity estimation, based on the phase difference technique, is presented. The proposed technique combines the robustness of the matching method with the sub-pixel accuracy of the phase difference approach. A matching between the phases of the left and right signals is introduced in order to allow the phase difference method to work in a reduced disparity range. In this framework, a new criterion to detect signal singularities is proposed. The presented test cases show that the performance of the proposed technique in terms of accuracy and density of the disparity estimates has greatly improved.

Unconstrained stereoscopic matching of lines

The computation of horizontal binocular disparities used in stereoscopic depth perception depends upon the identification of corresponding features in the two retinal images. In principle, binocular matching is a two-dimensional problem that considers matches in all possible meridians. Normally, constraints such as end points or crossing points limit the direction and magnitude of matches. If matching is unconstrained, such as is the case with long lines, it is completely ambiguous. Under these conditions the default match will be determined by the operating range, or upper disparity limit, of matchable vertical and horizontal disparities. We computed the operating range of vertical matches for stereoscopic depth as a function of line orientation. Our results suggest that the two-dimensional operating range is anisotropic for vertical and horizontal disparity and that unconstrained matches are not based upon either epipolar geometry or nearest neighbor constraints, but rather the mean of disparity estimates within the operating range for binocular matches. This operating range can be extended vertically when matches are constrained by image primitives.

Self-Describing Data Transfer Methodology for Intelligent Transportation Systems Applications

The wide variety of remote sensors used in intelligent transportation systems (ITS) applications (loops, probe vehicles, radar, cameras, etc.) has created a need for general methods by which data can be shared among agencies and users who own disparate computer systems. In this paper a methodology is presented that demonstrates that it is possible to create, encode, and decode a self-describing data (SDD) stream using existing data description language standards, parsers to enforce language compliance, a simple content language that flows out of the data description language, and architecture neutral encoders and decoders based on Abstract Syntax Notation One (ASN.1). Further, it documents a reference implementation of an SDD receiver.

How the owl resolves auditory coding ambiguity.

The barn owl (Tyto alba) uses interaural time difference (ITD) cues to localize sounds in the horizontal plane. Low-order binaural auditory neurons with sharp frequency tuning act as narrow-band coincidence detectors; such neurons respond equally well to sounds with a particular ITD and its phase equivalents and are said to be phase ambiguous. Higher-order neurons with broad frequency tuning are unambiguously selective for single ITDs in response to broad-band sounds and show little or no response to phase equivalents. Selectivity for single ITDs is thought to arise from the convergence of parallel, narrow-band frequency channels that originate in the cochlea. ITD tuning to variable bandwidth stimuli was measured in higher-order neurons of the owl's inferior colliculus to examine the rules that govern the relationship between frequency channel convergence and the resolution of phase ambiguity. Ambiguity decreased as stimulus bandwidth increased, reaching a minimum at 2-3 kHz. Two independent mechanisms appear to contribute to the elimination of ambiguity: one suppressive and one facilitative. The integration of information carried by parallel, distributed processing channels is a common theme of sensory processing that spans both modality and species boundaries. The principles underlying the resolution of phase ambiguity and frequency channel convergence in the owl may have implications for other sensory systems, such as electrolocation in electric fish and the computation of binocular disparity in the avian and mammalian visual systems.