Toward Perceptually-Consistent Stereo: A Scanline Study

  title={Toward Perceptually-Consistent Stereo: A Scanline Study},
  author={Jialiang Wang and Daniel Glasner and Todd E. Zickler},
  journal={2017 IEEE International Conference on Computer Vision (ICCV)},
Two types of information exist in a stereo pair: correlation (matching) and decorrelation (half-occlusion). Vision science has shown that both types of information are used in the visual cortex, and that people can perceive depth even when correlation cues are absent or very weak, a capability that remains absent from most computational stereo systems. As a step toward stereo algorithms that are more consistent with these perceptual phenomena, we re-examine the topic of scanline stereo as… 

Figures from this paper

Level Set Binocular Stereo with Occlusions

An energy and level-set optimizer that improves boundaries by encoding occlusion geometry is introduced that applies to two-layer, figure-ground scenes, and it can be implemented cooperatively using messages that pass predominantly between parents and children in an undecimated hierarchy of multiscale image patches.

Layered Stereo by Cooperative Grouping with Occlusion

This work introduces a layered approach to stereo that explicitly incorporates occlusions, and demonstrates the model's ability to localize boundaries between figure and ground in a wide variety of cases, including images from Middlebury and Falling Things datasets.

Local Detection of Stereo Occlusion Boundaries

The local detector for stereo occlusion boundaries is introduced based on a simple feedforward network with relatively small receptive fields and produces better boundaries than many other stereo methods, even without incorporating explicit stereo matching, top-down contextual cues, or single-image boundary cues.

Level Set Stereo For Cooperative Grouping With Occlusion

An energy and level-set optimizer is introduced that improves boundaries by encoding the essential geometry of occlusions: the spatial extent of an occlusion must equal the amplitude of the disparity jump that causes it.



The role of partial occlusion in stereopsis

MODELS of stereopsis typically assume that all the information about stereoscopic depth is contained in the disparity field, that is, the positional differences of image features that arise from

Occlusions and binocular stereo

A theory for stereo is described based on the Bayesian approach, using adaptive windows and a prior weak smoothness constraint, which incorporates occlusion, and it is shown that occlusions can help stereo computation by providing cues for depth discontinuities.

Toward a general theory of stereopsis: binocular matching, occluding contours, and fusion.

The authors hypothesize the existence of receptive field structures that sense the local structure of stereoscopic occlusion relationships to account for these findings and present a simple theoretical framework in which fusion, stereopsis, and Occlusion are unified.

Low-level vision by consensus in a spatial hierarchy of regions

A multi-scale framework for low-level vision, where the goal is estimating physical scene values from image data-such as depth from stereo image pairs-is introduced, which produces a distributional scene representation that is appropriate for combining with higher-level reasoning and other low- level cues.

Perception of multiple transparent planes in stereo vision

The result shows that stereo-matching is not necessarily unique—a given point of the image in one eye may be matched simultaneously to more than one point in the other eye, each match defining a different depth plane.

Shape and the Stereo Correspondence Problem

We examine the implications of shape on the process of finding dense correspondence and half-occlusions for a stereo pair of images. The desired property of the disparity map is that it should be a

The Role of Monocular Regions in Stereoscopic Displays

The basis of the present study was a prediction that the presence of texture within these regions would facilitate rather than retard stereoscopic processing, which follows from a hypothesis that stereoscopicprocessing is initially located at disparity discontinuities.

A Bayesian treatment of the stereo correspondence problem using half-occluded regions

  • P. BelhumeurD. Mumford
  • Computer Science
    Proceedings 1992 IEEE Computer Society Conference on Computer Vision and Pattern Recognition
  • 1992
The authors first derive a measure for goodness of fit and a prior based on a simplified model of objects in space, which leads to an energy functional depending both on the depth as measured from a central cyclopean eye and on the regions of points occluded from the left and right eye perspectives.

Detecting Binocular Half-Occlusions: Empirical Comparisons of Five Approaches

This paper considers five methods that explicitly detect half-occlusion points and quantitatively and qualitatively compare these methods under a variety of conditions.

Solving da Vinci stereopsis with depth-edge-selective V2 cells