Maimone, M., "Modeling foreshortening in stereo vision using local spatial frequency," Proceedings 1995.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....the template size is small or the true object s surface is large, the plane approximates the surface. The derived matching j kl i T T f f = but with different template shapes j T T s s . When compared to the work of Maimone and Shafer using Local Spatial Frequency representation [14], the rotation of 3D template is more intuitive; nevertheless, it is also more time consuming. Figure 4 shows four templates of four virtual voxels created using the methods discussed. Observe that the size of the templates appears different in different cameras. These templates size is ....

M.W. Maimone and S.A. Shafer, "Modeling Foreshortening in Stereo Vision using Local Spatial Frequency," Proceedings of 1995.

....projection of an object varies inversely with the distance of that object from the center of projection. Specifically, under perspective projection, parallel lines do not project to parallel lines and length ratio cannot be kept invariant. Previous work on modeling foreshortening (e.g. [4,17]) reveals that foreshortening can worsen correspondence results because 3D space is mapped nonlinearly to 2D image planes. Foreshortening in perspective stereo also results in highly biased scene reconstruction. Even if the correspondences between two perspective images are available with high ....

....projection cameras in real world. In fact, parallel projection stereo does not necessarily rely on the practical availability of parallel projection cameras in real world. In this paper, we transform a dense collection of perspective images, specifically 3D or 4D uniform light field dataset [3,4,9], into a set of parallel projection images. Light field data can be acquired by moving a camera uniformly along a line or on a planar grid. 1.1 Related work The work described in this paper is closely related to current research trends in multi perspective projection for image based ....

. Mark W. Maimone and Steven A. Shafer, Modeling the Foreshortening in Stereo Vision using Local Spatial Frequency, CMU Computer Science Technical Report TR95. Carnegie Mellon University, January 1995 (32 pages)

....respect to the derivatives of disparity, not to the surface orientation. Thereby, it is rather complex to utilize more than two images for a robust depth recovery because the disparities of different image pairs are not identical, nor are the derivatives of the disparities consequently. Mainome[14] tackled the deformation problem in phase based stereo, but the scheme still needs the brute force search over the surface orientation in addition to the depth search. Also in motion estimation or feature tracking, several researchers addressed the deformation problem of the intensity patterns. ....

M.W.Maimone and S.A.Shafer. Modeling foreshortening in stereo vision using local spatial frequency. In Proc.IROS, pages 519--524, 1995. 366 British Machine Vision Conference

....using simple disparity estimates from phase differences, and analyzed the errors which accrue when using such a method. Jenkin and Jepson (1994) used a cross correlation method on band pass filtered images to find the phase differences, and compared such a method to other stereopsis methods. Maimone and Shafer (1995) implemented a basic phase based algorithm and then extended it to account for the effects of perspective foreshortening, a problem which plagues stereo matching systems. Jones and Malik (1992) also implemented an algorithm which attempts to deal with the perspective foreshortening problem. The ....

....in the scene will generally give rise to an expansion of that surface in one image relative to the other, and hence to different spatial frequencies in the two images. For stereo cameras with a baseline B and a surface at a distance Z viewed at an angle , the expansion is given by : 1 B tan =Z (Maimone 11 and Shafer, 1995). The expansion parameter fi to be used in (26) can thus be related to the maximum allowable surface angle. However, this does not imply that surfaces at all viewing angles will give accurate results. It only means that we are prepared to accept as accurate results from viewing angles up to a ....

[Article contains additional citation context not shown here]

Maimone, M. W. and Shafer, S. A. (1995). Modeling foreshortening in stereo vision using local spatial frequency. Technical Report CMU-CS-95-104, Carnegie Mellon University, Pittsburgh, PA 15213.

....is also useful for discerning and describing any theoretical limitations of the method, e.g. the points at which its assumptions break down. 2. 1 Example 1: Phase difference as disparity The use of phase difference as disparity lies at the heart of many phase based stereo algorithms [6] 7] 5] [8]. But unless one is already quite familiar with the frequency domain, the name itself inspires fear. Suppose one remained uncertain of the underlying principles; how could you convince yourself that the technique really works By considering the simplest possible examples according to the ....

M. W. Maimone and S. A. Shafer, "Modeling foreshortening in stereo vision using local spatial frequency," in International Robotics and Systems Conference (IROS), pp. 519--524, IEEE Computer Society Press, Aug. 1995. http://www.ius.cs.cmu.edu/project/cil/fore/tr.html.

No context found.

Maimone, M., "Modeling foreshortening in stereo vision using local spatial frequency," Proceedings 1995.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC