A. Blake, A. Zisserman, and G. Knowles. Surface descriptions from stereo and shading. IVC, 3(4):183--191, 1985.  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... on the occluding contour are the maximum (giving the edge element in c) and the intersection of the contour line with the self shadow edge (giving the edge elements in d) planar, cylindrical, or hyperbolic, provides one important intrinsic surface feature (see also [1] 28] 30] 20] [3] and [29] With this classification of parts as areas of the same sign of Gaussian curvature, part boundaries within an object are located on parabolic lines. The parts produced by this segmentation are often qualitatively similar to the parts produced by a generalized cylindersbased scheme. I ....

A. Blake, A. Zisserman, and G. Knowles. Surface descriptions from stereo and shading. In B. K. P. Horn and M. J. Brooks, editors, Shape from shading, pages 29--52. The MIT press, Cambridge, MA, 1989.

....depth information from that along the boundary DA. Shapefrom shading provides (relative) depth information over the region A, from those along the boundary OA. Both modules are complementary in nature, and thus combining the methods is an interesting research goal (Ikeuchi [8,9] Blake et al. [1]) Known reflectance map Our shape from shading algorithm requires a known exact reflectance map, that is, known reflectance characteristics, and a known light source direction. 94 K. lkeuchi It is rather unusual to know the exact reflectance characteristics of an object and the exact direction ....

A. Blake, A. Zisserman and G. Knowles, Surface descriptions from stereo and shading, Image Vis. Cornput. 3-4) (1985) 183-191.

....is not possible without additional constraints, due to the number of unknowns and nonlinearity of the system. Obtaining a unique solution for SFS for constant albedo approximation was given in [8] by propagating the known solutions at singular points to the image. Similar results are also given in [2] for boundary conditions. However, in [14] it was shown that methods that guarantee unique solutions by propagation [1] are out performed by minimization methods, which do not guarantee a unique solution. In this section, we will describe the symmetry constraint for defining an albedo free ....

A. Blake and A. Zisserman. Surface descriptions from stereo and shading. Image and Vision Computing Journal, 3(2):183-191, 1992.

....used to estimate accurate depth maps of a scene. Other contributions consider the fusion of 2D and 3D data in the form of intensity images and stereo or range data [1, 9] or only 3D data acquired from stereo and range data [7, 8] Methods that merge data from stereo and shape from shading include [2, 3, 5]. Most of the methods that fuse depth information are designed to work at a point or pixel level, whereas those methods that include perceptual grouping in 2D images are mostly directed towards fusing higher level primitives [10] or more specifically, at labelling primitives such as contours or ....

A. Blake, A. Zisserman, and G. Knowels. Surface description from stereo and shading. In B. Horn and M. Brooks, editors, Shape from Shading, Artificial Intelligence Series, chapter 2, pages 29--52. The MIT Press, Cambridge, Aug. 1989.

....occluding contours is available. Their method iteratively computes the surface orientation at the remaining locations by propogating the depth at the occluding contours. The contour depth map computed by the feature based stereo can be used for this purpose. In fact, Blake, Zisserman and Knowles [3] have shown that if the boundary information (depth) is available at the occluding contours then shape from shading converges to a unique solution. In a recent shape from shading algorithm reported by Leclerc and Bobik [11] which uses the conjugate gradient method for minimizing the cost ....

Blake, A., Zisserman, A. and Knowles, G. Surface Descriptions from Stereo and Shading. Image and Vision Computing, Vol. 3, No. 4, pp. 183-191, 1985.

....used to estimate accurate depth maps of a scene. Other contributions consider the fusion of 2D and 3D data in the form of intensity images and stereo or range data [1, 9] or only 3D data acquired from stereo and range data [7, 8] Methods that merge data from stereo and shape from shading include [2, 3, 5]. Most of the methods that fuse depth information are designed to work at a point or pixel level, whereas those methods that include perceptual grouping in 2D images are mostly directed towards fusing higher level primitives [10] or more specifically, at labelling primitives such as contours or ....

A. Blake, A. Zisserman, and G. Knowels. Surface description from stereo and shading. In B. Horn and M. Brooks, editors, Shape from Shading, Artificial Intelligence Series, chapter 2, pages 29--52. The MIT Press, Cambridge, Aug. 1989.

....from an image involves an inference process. There is a broad literature on a set of techniques for inferring such information based on explicit reconstruction of the 3# shape. These include shape from shading [1, 2, 3, 4] shape from texture [5, 6, 7, 8] shape from motion [9] shape from stereo [10, 11], and many other similar processes collectively referred to as shape from X. While these algorithms are gaining mathematical rigor [9] they still lack the robustness that is necessary to be useful in a general environment. Other approaches include extraction of qualitative3# information from ....

A. Blake, A. Zisserman, and G. Knowles. Surface descriptions from stereo and shading. In B. K. P. Horn and M. J. Brooks, editors, Shape from Shading, chapter 2, pages 29-52. Massachusetts Institute of Technology, Cambridge, Massachusetts, 1989.

....the methods mentioned above. Previous attempts to integrate photometric and geometric constraints were based on using each of the methods in different image regions. In some studies the results computed by say, a stereo algorithm, were transfered to a photometric shape recovery algorithm (e.g. [2, 4, 7, 15]) To the best of our knowledge, our scheme is the first to propose integration in the strong sense. The integration of the photometric and geometric constraints allow us not only to compute the selfcorrespondence function, but also to recover the shape of the object based on geometric ....

A. Blake, A. Zisserman, and G. Knowles. Surface description from stereo and shading.Image Vision Computation, 3(4):183# 191, 1985.

....a normalized cross correlation criterion, the underlying assumption being that of brightness constancy of corresponding points. dence wherever the brightness gradient is non zero. Issues concerning the fusion of shading cues with parallax cues have been discussed in several works, including [1, 31, 18, 34, 20, 7, 3, 30]. 2 A generative model Let S be a surface in space of class at least C 1 ; we indicate the tangent plane to the surface at a point P by T and the inward unit normal vector by N . At each P 2 S we can construct a reference frame with origin at P , e 3 axis parallel to the normal vector N and ....

A. Blake, A. Zisserman, and G. Knowles. Surface descriptions from stereo and shading. Image and vision computing, 3(4):183-191, 1985.

....wherever the brightness gradient is non zero, or when it is greater than a threshold in the presence of noise, thus avoiding an ill posed formulation of the reconstruction problem. Issues concerning the fusion of shading cues with parallax cues have been discussed in several works, including [1, 29, 17, 32, 19, 7, 3, 28]. To our knowledge, the merging of multi view geometry with shading in a variational framework is novel. Also, the re visitation of the correspondence problem and the de nition of correspondence sets is novel, and so is the fast level set implementation of the stereoscopic shading algorithm that ....

A. Blake, A. Zisserman, and G. Knowles. Surface descriptions from stereo and shading. Image and vision computing, 3(4):183-191, 1985.

....effects of motion parallax and changes in radiance from an object s surface (dynamic shading) Discussion of these combined effects has been limited in previous work on qualitative shape analysis. Prior work has focused either on motion parallax [9, 1, 15, 14, 6, 16, 5] or dynamic shading [10, 11, 4, 17, 3, 2], but not both. Although Verri and Poggio [15] considered how photometric effects influ The support of the National Science Foundation under Grant No. IRI 9022608 is gratefully acknowledged. We also thank J. Barron and S. Beauchemin for providing the optical flow computation software. The ....

....are important because motion parallax creates difficulties in measuring changes in shading. Koenderink and van Doorn [10, 11] present their results exclusively in terms of object surface radiance, thus avoiding the need to discuss the effects of motion parallax on image irradiance. Blake et al. [4] presented a technique based on Koenderink and van Doorn s results, which provided a means of locating parabolic curves. However, they avoided the problem of object motion by moving the light source instead of the object. Closely related to our work is that of Blake and Bulthoff [2] who used ....

A. Blake, A. Zisserman, and G. Knowles. Surface descriptions from stereo and shading. Image and Vision Computing, 3:183--191, 1985.

....solved directly when z is known. 3.3 Existence and Uniqueness Whether we formulate shape from shading as a problem of differential equations or as one of minimization, we should consider the issue of the existence of a solution and whether a solution is unique. Results in the continuous domain (Blake et al. 1985, Bruss 1982) have shown a unique solution to the shape from shading problem under the restrictive conditions that the light source direction is equal to the viewing direction, boundary conditions are completely specified, and that the data are noise free. Recently, Oliensis (1991) showed that if ....

Blake, A., Zisserman, A., and Knowles, G. (1985). Surface descriptions from stereo and shading. Image Vision Comput., 3(4), 183--191.

....image, yielding the 3D shape. Previous attempts to integrate photometric and geometric constraints were based on using each of the methods in different image regions. In some studies the results computed by, say, a stereo algorithm, were transfered to a photometric shape recovery algorithm (e.g. [1, 3, 6, 14]) To the best of our knowledge, our scheme is the first to propose integration in the strong sense. The integration of the photometric and geometric constraints allow us not only to compute the self correspondence function, but also to recover the shape of the object based on geometric ....

A. Blake, A. Zisserman, and G. Knowles. Surface description from stereo and shading. Image Vision Computation, 3(4):183--191, 1985.

....albedo. Consequently, the natural choice for the monocular information source is shading, while intensity is the natural choice for the image feature used in multi image correspondence. Not only are these the natural choices given a Lambertian reflectance model, they are also complementary (Blake et al. 1985, Leclerc and Bobick 1991) intensity correlation is most accurate wherever the input images are highly textured, whereas shading is most accurate where the input images are untextured. The reconstruction method is to minimize an objective function whose components depend on the input images and ....

....to deriving an initial estimate for the mesh and Szeliski and Tonnesen s approach (1992) are different from ours and this is an important topic for future research. Of special relevance to this paper is research in combining stereo and shape from shading. Using 2 1 2 D representations, Blake et al. 1985) is the earliest reference we are aware of that discusses the complementary nature of stereo and shape from shading, but meaningful experimental results are not provided. Leclerc and Bobick (1991) discuss the integration of stereo and shape from shading, but their implementation uses stereo only ....

Blake, A., Zisserman, A., and Knowles, G. (1985). Surface descriptions from stereo and shading. Image Vision Computation, 3(4), 183--191.

....hyperbolic points can be identified simply by observing the rotational change of the image intensity gradient. It has already been shown in previous work that in the Lambertian case (Koenderink and Van Doorn ) 13] and in the generalized case of all reflectance maps (Blake Zisserman and Knowles) [5] that the direction of the image intensity gradient at parabolic points is invariant to all illumination conditions, a result that falls out as a specific instance in our formulation. Yuille [21] has shown that the invariant direction of the image intensity gradient is perpendicular to the ....

....our formulation. Yuille [21] has shown that the invariant direction of the image intensity gradient is perpendicular to the direction of zero curvature. This was also noticed in [17] A method for finding parabolic points on a smooth object surface was suggested in Blake, Zisserman, and, Knowles [5] by selecting points where the orientation of the image intensity gradient remains the same. In our technique parabolic points are equivalently identified by the specific instance where the image intensity gradient exhibits zero rotational change. Unfortunately Lambertian reflectance under varying ....

A. Blake, A. Zisserman, and G. Knowles. Surface descriptions from stereo and shading. Image and Vision Computing, 3(4):183--191, 1985.

....and the relationship of this approach to previous work. Here, we focus on how an additional image cue (silhouette edges) and previously computed 3 D data are incorporated into our reconstruction process. Combining these different sources of information is not a new idea in itself. For example, Blake et al. 1985 ] is the earliest reference we are aware of that discusses the complementary nature of stereo and shape from shading. Both Cryer et al. 1992 ] and Heipke et al. 1992 ] have proposed algorithms to combine shape from shading and stereo, while Liedtke et al. 1991 ] first uses silhouettes to ....

A. Blake, A. Zisserman, and G. Knowles. Surface descriptions from stereo and shading. Image Vision Computation, 3(4):183--191, 1985.

....occluding contours is available. Their method iteratively computes the surface orientation at the remaining locations by propogating the depth at the occluding contours. The contour depth map computed by the feature based stereo can be used for this purpose. In fact, Blake, Zisserman and Knowles [3] have shown that if the boundary information (depth) is available at the occluding contours then shape from shading converges to a unique solution. In a recent shape from shading algorithm reported by Leclerc and Bobik [11] which uses the conjugate gradient method for minimizing the cost ....

Blake, A., Zisserman, A. and Knowles, G. Surface Descriptions from Stereo and Shading. Image and Vision Computing, Vol. 3, No. 4, pp. 183-191, 1985.

No context found.

A. Blake, A. Zisserman, and G. Knowles. Surface description from stereo and shading. IVC, 3(4):183--191, 1985.

No context found.

A. Blake, A. Zisserman, and G. Knowles. Surface description from stereo and shading. IVC, 3(4):183--191, 1985.

No context found.

A. Blake, A. Zisserman, and G. Knowles. Surface descriptions from stereo and shading. IVC, 3(4):183--191, 1985.

No context found.

A. Blake, A. Zisserman, and G. Knowles. Surface description from stereo and shading. Image and Vision Computing, 3(4):183--191, 1985.

No context found.

A. BLAKE,A.ZISSERMAN, AND G. KNOWELS, Surface description from stereo and shading. In B. Horn and M. Brooks, eds., Shape from Shading, Artificial Intelligence Series, chap. 2, pp. 29--52, The MIT Press, Cambridge, August 1989.

No context found.

A. Blake, A. Zisserman, and G. Knowles, "Surface Descriptions from Stereo and Shading", Image and Vision Computing, Vol. 3, No. 4, 1985, pp. 183-191.

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