Kanade, T. 1981. Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence, 17:409--460.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....have high time complexity. Symmetry in 3D is usually approached from the 2D projections of the 3D objects. The symmetry can be determined directly from the projections [83, 71, 38, 73, 82, 117, 72, 34] or the 3D symmetric object can be reconstructed from its projection and 3D symmetry determined [53, 97, 114, 106, 70, 33]. Additional approaches to detection of 3D symmetry involve deformable contours [99] tomography [27] structure from motion techniques [80, 67, 88] and physical mirrors [74, 51, 1, 54] In most of the above mentioned techniques, symmetry is treated as a binary feature: either it exists or it ....

....the p Gamma q space at angles corresponding to the skew symmetry axis in the image (a) angles ff; fi) All possible orientations of a surface can be represented as a 2D gradient space p Gamma q where p denotes the gradient in the x coordinate direction and q in the y coordinate direction. In [53], it is shown that for a given skew symmetry axis, the possible orientations of the original mirror symmetric planar figure is constrained to lie on a hyperbola in the p Gamma q gradient space. The asymptotes of the hyperbola are shown to be orthogonal to lines drawn in the p Gamma q space at ....

[Article contains additional citation context not shown here]

T. Kanade. Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence, 17:409--460, 1981.

....contains all symmetry pair midpoints. Methods to exploit symmetry and infer constraints on the 3D reconstructed shape of the object scene have a long history in Computer Vision. Based on the assumption that a skew symmetry depicts a real symmetry viewed from some unknown angle (Kanade [5]) a large number of approaches rely on the analysis of skewed symmetries in images to infer constraints on the 3 D geometry of depicted objects. In most cases, scaled) orthographic projection is assumed (see e.g. 7] 2] A more limited number of studies deal with symmetry observed under ....

Kanade T. Recovery of the Three-Dimensional Shape of an Object from a Single View. Artificial Intelligence, vol. 17, 1981, pp. 409-460.

....it is an invariant projective property, and for trihedral drawings it can be checked with a pencil and an unmarked ruler alone. 1. Introduction Emulating the human performance in interpreting drawings of polyhedra has been one of the goals of Computer Vision along the past three decades [2, 13, 12, 3, 1]. A usual motivation behind the extensive work done in the area is helping to elucidate why humans are able to reject impossible figures , and recover 3D shapes from correct ones (fig. 1) despite the reduced information they offer, without textures on the surfaces, illumination patterns, or ....

T. Kanade. Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence, 17:409--460, 1981.

....have extremely useful information for delineating object boundaries and for discriminating physical phenomena around edges. Although there has been limited research on this topic, there are a few related works. Kanade proposed a method for using edge profiles in making a pair or group of edges [1]. Marik, et. al, proposed edge discrimination based on color distribution [3] Maxwell, et. al, proposed a method to utilize an edge profile for image segmentation [4, 5] They proved the utility of gray or color distribution around an edge. We propose a novel method for using this information to ....

T. Kanade. Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence, Vol.17, pages 409--460, 1981.

....density gradient as the primary basis for surface perception by humans. Formalizing the shape from texture problem requires modeling the image forming system. Mainly two kinds of projections have been commonly used: the orthographic projection and the perspective projection. Kender [7] Kanade [8], and Witkin [10] explored the domain under the orthographic projection. Kender formalizes the relationship between local surface orien tation and two perpendicular texture elements assumed to have the same Artificial Intelligence 22 (19841 49 75 0004 3702 84 3.00 1984, Elsevier Science ....

....orien tation and two perpendicular texture elements assumed to have the same Artificial Intelligence 22 (19841 49 75 0004 3702 84 3.00 1984, Elsevier Science Publishers B.V. North Holland) 50 K. KEUCHI length [7] Kanade proposes to use skewed symmetry [9] to recover local surface orientation [8]. The angle between a skewed symmetry direction and the opposite direction can be a constraint to reduce possible surface orien tations. Witkin determines the most likely surface orientation, based on the observed angular distribution of the texture assuming that the texture elements have random ....

[Article contains additional citation context not shown here]

Kanade, T., Recovery of the three-dimensional shape of an object from a single view, CMU-CS-79-153, Dept. Computer Science, CMU, 1979.

....axes. This assumption made the algorithm very fast; real time performance was achieved. 3. 3 Skewed symmetries and line segment pairs for object recognition Extracting skew symmetries has been another approach to make shape interpretations for deriving information about 3 D surface orientations [14, 16, 24, 34]. Most of these methods are tested only for line drawings [14, 24] Ponce [34] used gray value images for the experiments, however, the contours analyzed were closed and they were selected by hand. In [16] the construction of a symmetry analyzer is described, but occlusions are not allowed and the ....

....was achieved. 3. 3 Skewed symmetries and line segment pairs for object recognition Extracting skew symmetries has been another approach to make shape interpretations for deriving information about 3 D surface orientations [14, 16, 24, 34] Most of these methods are tested only for line drawings [14, 24]. Ponce [34] used gray value images for the experiments, however, the contours analyzed were closed and they were selected by hand. In [16] the construction of a symmetry analyzer is described, but occlusions are not allowed and the segmentation into object and background is assumed to be already ....

T. Kanade, Recovery of three-dimensional shape of an object from a single view, Artificial Intelligence 17, 1981, 409-460.

.... from images of planar objects, or from a single silhouette of 3D objects, viewed in a fronto parallel plane [3, 4, 26] The methods developed for those cases can not be applied if the viewpoint is not fronto parallel since, for a planar object, reflectional symmetries are then skewed by imaging [13, 17, 29]. For a 3D smooth object additional distortions occur the contour generator moves on the surface as viewing position changes. In this case it is not a fixed space curve which is projected, and the image profile can change radically with viewpoint, defeating any simple application of skewed ....

Kanade T., `Recovery of Three Dimensional Shape of an Object from a Single View'. Artificial Intelligence, vol. 17, nos. 1--3, pp. 409--460, 1981.

....shape has two axes of symmetry but under perspective viewing the image shape is not symmetrical 6. 6Symmetry is also lost under orthographic projection, but the concept of skewed symmetry can be defined which enables the recovery of symmetrical figures and constrains the pose of the figure planes [166]. 473 A perspective drawing with major vanishing points indicated. Plan View ,About c Perspective View Figure 23. Symmetry is lost under perspective projection. The dashed line indicates a reflection of the perspectively transformed shape about the line c. 474 View Light Source ....

Kanade, T., Recovery of the Three-Dimensional Shape of an Object from a Single View, Artificial Intelligence, 17, 409-460, 1981.

....to be taken to be able to handle more general types of objects. A central notion in 3 D inference is symmetry. Kanade observed that the ability to infer the 3 D of a volume observed in a picture relies on the assumption that a skew symmetry depicts a real symmetry viewed from some unknown angle [8]. A property that numerous objects share is indeed bilateral symmetry in 3 D. The GCs we considered so far are simply special cases of bilateral symmetric objects. Two true edges that are bilateral symmetric in 3 D project into two image edges that define a skew symmetry (assuming non accidental ....

Kanade T., "Recovery of the Three-Dimensional Shape of an Object from a Single View". In Artificial Intelligence, vol. 17, 1981, pp. 409-460.

....in parallel with the stereo work, there are work that recovers shapes from a single image. Questions are, what are the limitations of such monocular methods, and precisely howmuch more can our stereo method offer Early monocular work was focussed on the analysis of line drawings of polyhedra [12, 9, 15, 13]. There have also been some primitive attempts to handle curved surfaces such as [1, 22, 27, 11] Ponce et al. 18] examined invariant properties of 3 D shapes in their 2 D projections, and showed that the shape reconstruction problem can be simplified in the case where the objects viewed are GCs. ....

T. Kanade. Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence, 17:409--460, 1981.

....plane faced objects from the trihedral or Origami world, for which simpler line labeling procedures can be applied. Single projection line drawing interpretation has been treated qualitatively by Huffman [6] and Clowes [7] and quantitatively using line labeling techniques by, for example, Kanade [8] and Sugihara [9] According to these techniques, an edge in a drawing can represent one of three states: concave face connection, convex face connection, or occluding face edge. A consistent set of line labels is searched for, using a library of possible line junction configurations. Many such ....

....highlighted, b) natural solution (c) a second valid but unlikely solution. Selecting the most plausible solution is a matter of determining what makes solution 13(b) more apparent to the human eye. This determination is based on perception of image regularities. The concept of image regularities [8] specifies that certain geometrical relationships between entities in the projection plane suggest spatial relationships. For example, the fact that two lines are parallel in the projection plane usually implies that they are also spatially parallel. The solution in Fig. 13(c) is not plausible ....

T. Kanade, "Recovery of the three-dimensional shape of an object from a single view," Artificial Intelligence, Vol. 17, pp. 409-460, 1980.

....it is an invariant projective property, and for trihedral drawings it can be checked with a pencil and an unmarked ruler alone. 1. Introduction Emulating the human performance in interpreting drawings of polyhedra has been one of the goals of Computer Vision along the past three decades [2, 13, 12, 3, 1]. A usual motivation behind the extensive work done in the area is helping to elucidate why humans are able to reject impossible figures , and recover 3D shapes from correct ones (fig. 1) despite the reduced information they offer, without textures on the surfaces, illumination patterns, or ....

T. Kanade. Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence, 17:409--460, 1981.

.... processing errors the difference between N and N is ignored except for its use in rotations, and rotation axes are identified at a processing stage [17] which only considers fully visible faces) It is possible to attempt to refine these normal direction estimates using skewed symmetry [5]. We have found no benefit in doing this. Even without using skewed symmetry, the estimates are accurate enough to be used as input to 2 the remaining stages of the process. The effect of improving them would be to reduce the time taken by the iterative optimisation. Since skewed symmetry ....

....We found amoeba to be even slower than the geometry specific optimisation described above, and although it remains a possibility to which we might return, it is not our current recommendation. In both our current system and the alternative using amoeba, we have tried using skewed symmetry [5] to produce a better initial estimate of face normals, in the hope that this would help the optimisation process to converge more quickly. From Kanade [5] skewed symmetry is a method for calculating the normal of a face given two line (axes) on the corresponding region which are believed to be ....

[Article contains additional citation context not shown here]

T.Kanade, Recovery of the Three-Dimensional Shape of an Object from a Single View, Artificial Intelligence 17 409--460, 1981.

....For these to function correctly, it is desirable to have a topologically correctcompletion of the visible part of the object with added depth information. Accurate geometry is not required, but the depth ordering of the adjacent visible vertices must be correct (for example: skewed symmetry [10] produces multiple interpretations for the projection used to draw the sketch, and selection of the correct one is based on knowledge of the relative depths) We have developed a quick method which produces numerical values which give correct depth ordering in all of the cases we have tested. ....

....and lines joining these pairs are close to parallel with one another and centred on the mirror axis. The angle of the axis, and the average angle of the lines joining sequential pairs, are recorded they are the input angles (#, #) required by skewed symmetry for estimating face normals [10], an option which, although not part of our current system, has been used with success elsewhere [6] The figure of merit for region mirror symmetry is the product of the figures of merit for the ratios of distances either side of the line of matched vertices and the figures of merit for ....

T.Kanade, Recovery of the Three-Dimensional Shape of an Object from a Single View, Artificial Intelligence 17, 409--460, 1981.

....digit recognition 1. Introduction The detection of symmetries is an essential step when inferring shapes from contours [1] Particularly, skewed symmetries and parallel symmetries have been shown to be useful in recovering the 3D structure of various types of objects from their 2D contours [2,3]. Results from the study of perceptual organization in Ref. 4] also showed that primitive grouping in our vision system is done on proximity, collinearity, curvilinearity, parallelism, symmetry, closure and repetitive patterns. Symmetries (including parallelism) have played important roles in ....

T. Kanade, Recovery of the three-dimensional shape of an object from a single view, Arti"cial Intell. 17 (1981) 409}460.

....convenience, we call the perspective projection of 3 D bilateral symmetry perspective symmetry. Symmetry has been an active research area in the computer vision community for the past two decades or so. A plethora of approaches have been reported in the context of skewed symmetry analysis (e.g. [3 6]) These approaches assume (scaled) orthographic projection, and typically focus on the detection of the symmetry axis of 2 D bilateral symmetric shapes on the image plane. In contrast, work on perspective symmetry analysis has been very limited indeed. The best known published work in the area ....

T. Kanade, Recovery of the Three-Dimensional Shape of an Object from a Single View, Artificial Intelligence, vol.17, 1981, pp.409-460.

....and general one. Consider, for example, the problem of obtaining a face on view of a planar curve (or a texture element) This problem has been studied extensively in the past and several approaches exist that hypothesize face on views, based on information from a single viewpoint (e.g. [42, 43]) We are currently investigating an approach similar to the one presented in this paper that enables the observer to change viewpoint in order to obtain a face on view of a planar curve. We are also trying to extend our results to perspective projection and are investigating possible uses of this ....

T. Kanade, "Recovery of the three-dimensional shape of an object from a single view," Artificial Intelligence, vol. 17, pp. 409--460, 1981.

.... symmetry as a constraint [31] 24] More recently, symmetry has been used to discriminate textures [9] and has been used in guiding robot grasping [7] Considerable work has been done on skewed symmetries including detection and exploiting them in the reconstruction of 3D structure [26] 27] [18], 14] 28] So far, symmetry has been treated as a binary feature: either it exists or it does not exist. The approach to symmetry as a continuous feature is novel, although several early studies have approached the question of measuring symmetry: In an early work, Grunbaum [15] reviews methods ....

T. Kanade. Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence, 17:409--460, 1981.

.... Many man made and natural objects display symmetry to an extent which, if robustly detected, could provide cues for scene segmentation and object recognition [19] Prior knowledge of 3D object symmetry has also been shown to aid in the recovery of 3D structure information from a single view [13, 20, 5]. Detecting 3D symmetry from a single view is a very difficult problem, however, because of the lack of constraints available. Consequently the bulk of the work published on bilateral symmetry detection falls roughly into two categories: i) planar symmetry from a single view, and (ii) 3D symmetry ....

....conditions, prior segmentation, and connectivity of features. Of the planar methods, those which detect bilaterally symmetric configurations of 2D image points have only limited applicability to images of 3D scenes. See [6] for a comprehensive bibliography. Skewed symmetry was defined by Kanade [13] as the constraints imposed by bilateral symmetry of planar figures on the geometry of an affine view of the figure. Thanks to David Murray for guidance and support to Paul Beardsley for kindly making his corner matcher software available and to the reviewers for their comments. The financial ....

[Article contains additional citation context not shown here]

T. Kanade. Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence, 17:409--460, 1981.

.... 1 Cotermination Julesz, 1971; Nakayama Shimojo, 1992 3,4,6 Circular Foster Wagemans, 1993; Lowe, 1985 1,3 Rectangular Gregory, 1970 7 Parallel Rock, 1983; Stevens, 1988 5,7 Rigid Body Motion Ullman, 1979; Wallach O Connell, 1953 1,8 Collinearity Westheimer McKee, 1977 2,5 Skew Symmetry Kanade, 1981; Leyton, 1992 Table 1: Some familiar non accidental relations often suggested as driving visual perceptual inferences (for acoustic analogs, see Bregman, 1990) The numbers in the third column refer respectively to: 1) Bennett, Hoffman Prakash, 1989; 2) Albert Hoffman, 1991; 3) Knill ....

Kanade, T. (1981) Recovery of the three-dimensional shape of an object from a single view.

....Second, we consider past efforts at making grouping systems efficient. Third, we describe the role that convexity has played in a variety of approaches to image understanding. The use of non accidental properties in grouping has been developed by Witkin and Tenenbaum[50] Binford[4] Kanade[31], Lowe[34] and Richards and Jepson[42] These authors argue that the visual system is sensitive to properties commonly produced by a single object or process, and rarely occurring at random. They show how this approach can explain a number of well known Gestalt grouping phenomena, including ....

Kanade, T., 1981, "Recovery of the Three-Dimensional Shape of an Object from a Single View," Artificial Intelligence, 17:409-460.

....in terms of a condition that has to be satisfied for two points to form a symmetry; the line connecting two points is called the line of symmetry. The middle point is the point in the middle of the line of symmetry. The axis of symmetry is formed by the union of middle points. Skew symmetry [198] is a symmetry where the lines connecting corresponding (symmetrical) points are perpendicular to the axis of symmetry. This is a mirror type symmetry. The notion of parallel symmetry was proposed by Ulupinar and Nevatia [70] This symmetry assumes that there is a continuous monotonic mapping ....

T. Kanade. Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence, 17:409--460, 1981.

....Parallel segments are also examined to detect other segments with proximity relations to their endpoints which were themselves parallel. Another higher level grouping is formed by checking pairs of proximity relations that are close to one another to see whether the four segments satisfy Kanade s [16] skewed symmetry relation (i.e. whether the segments could be the projection of segments that are bilaterally symmetric in three space) Since each of these compound structures is built from primitive relations that are themselves viewpoint invariant, the larger groupings also reflect properties ....

Kanade, Takeo, "Recovery of the three-dimensional shape of an object from a single view," Artificial Intelligence, 17 (1981), 409-460.

....can also be described by general affine transformations under the weak perspective assumption. Thus, the corresponding two curves of this symmetry have the same invariant signatures under general affine transformations. We must note the following properties: 1. The skewed symmetry proposed by Kanade (Kanade, 1981; Kanade and Kender, 1983) is a special case of this class of symmetry, where the rotational angle is equal to zero and the distortion can be described by a special affine transformation with determinant of 01. 2. Unlike the skewed symmetry of planar objects, 3D bilateral symmetry takes both ....

Kanade, T. 1981. Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence, 17:409--460.

....7, and the two weakly isotropic interpretations found by WISP are shown to the right. The first interpretation is the correct one where WISP as predicted 7 Experimental results 21 by Proposition 3 has recovered the rotational symmetry. The second interpretation has no symmetry whatsoever. Kanade [14] has discussed skewed symmetries, which are orthographic projections of planar reflectional symmetries. Kanade proposed the heuristic that skewed symmetries are interpreted as oriented real symmetries, and showed that this heuristic constrains the surface orientation to a one parameter family (a ....

....[5] 8.4 Weak isotropy and symmetries Symmetric shapes are important, because they often represent significant non accidental structure in the scene. This has led many researchers to look for specialized algorithms for symmetry detection [26] and in particular Kanade s skewed symmetries [14] have received considerable interest. We have shown that many symmetries are found automatically by the weak isotropy algorithm, namely skewed symmetries and all rotational symmetries of order three or higher. The result regarding rotational symmetries is particularly interesting, because we know ....

T. Kanade, "Recovery of the three-dimensional shape of an object from a single view", J. of Artificial Intelligence, vol. 17, pp. 409--460, 1981.

....when a sufficient set of point features of the object can be observed, and the position of these features on the object is known, the position of the object relative to the camera in world space can be reconstructed. For more detailed information about robust model based approaches, consult, e.g. (Kanade, 1981; Lamdan Wolfson, 1988; Gavrila Groen, 1992) An exemplar method is the following: when the camera is looking down to the scene consisting of a single object with a flat, horizontal surface, the observed area of the object is inversely proportional to the square root of the distance. 2. ....

Kanade, T. (1981). Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence, 17, 409--460.

....represents a temporally coherent structure in the ST cube, i.e. structures that result from an object or surface in the image undergoing similar image motion over long intervals of time. In order to recover 3D shape from a single image, Kanade used the heuristic of nonaccidental regularities [Kana81]: Regularities observable in the picture are not by accident, but are some projection of real regularities. to recover 3D shape from a single image. Similarly, we assume that similar image motion is the projection of similar scene motion. So a cluster of flow curves also corresponds to an ....

T. Kanade. Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence, 17:409--460, 1981.

....the edge of an object is occluded by another object. A few years later, a series of important discoveries about the combined topological and geometric properties of polyhedral scenes generated a formal basis for the interpretation of line drawings (Huffman [14] Clowes [5] Waltz [32] and Kanade [16]) These works demonstrated that a three dimensional interpretation of a polyhedral scene could be derived completely from the relationship between image vertices and edges the of the Fig. 1. A scene of a complexity typical of that which could be tackled by Guzman s SEE program (taken from ....

T. Kanade. Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence Journal, 17:409--460, 1981.

....that to identify the interpretation natural to a human observer it was necessary to use additional information including parallelism of lines, symmetry, and the hypothesis that any alignment or parallelism of lines detected in the image was not accidentally caused by the particular viewpoint. Kanade (1981) extended the Origami world method by using the heuristic that regularities in the image are not accidental, but are caused by real regularities in the scene. The regularities sought were parallelism and skewed symmetry, although it was noted that other kinds of regularity could also be used. The ....

Kanade, T., "Recovery of the Three-Dimensional Shape of an Object from a Single View," Artificial Intelligence 17, North-Holland, 1981, 409--460.

....one. Consider, for example, the problem of obtaining a face on view of a planar curve (or a texture element) This problem has been studied extensively in the past and several approaches exist that hypothesize planar views of the surface, based on information from a single viewpoint (e.g. [35, 36]) We are currently investigating an approach similar to the one presented in this paper that enables the observer to change viewpoint in order to obtain a face on view of the contour. Appendix: The Extent of Viewing Direction Adjustments Let us assume we have recovered the principal curvatures ....

T. Kanade, "Recovery of the three-dimensional shape of an object from a single view," Artificial Intelligence, vol. 17, pp. 409--460, 1981.

....from the image. Nevertheless, the heights of some of these structures can be determined using the rule that if two lines are aligned in the image, they are often aligned in 3 space. This rule has been used in other systems [20] and in fact is a restricted version of the parallel line rule [18] which states that parallel lines in the image often arise from INCREMENTAL RECONSTRUCTION OF 3D SCENES 3 [ 1 FIG. 21. Perspective view of 3D wire frames generated from Fig. 18 using the method of finding junctions arising from vertices lying on the ground plane. P5 P6 P9 PlO P8 Pll FIG. ....

Kanade, T., Recovery of the three-dimensional shape of an object from a single view, Artificial Intelligence 17(1-3) (1981) 409-460.

.... Intelligence 59 (1993) 95 101 95 Elsevier ARTINT 1002 From a real chair chair to a negative Takeo Kanade School of Computer Science, Carnegie Mellon University, Pittsburgh, PA 15213, USA 1. How it started In the fall of 1977, I was invited by Raj Reddy to spend a year at the Computer Science Department of Carnegie Mellon University as a ....

.... Intelligence 59 (1993) 95 101 95 Elsevier ARTINT 1002 From a real chair chair to a negative Takeo Kanade School of Computer Science, Carnegie Mellon University, Pittsburgh, PA 15213, USA 1. How it started In the fall of 1977, I was invited by Raj Reddy to spend a year at the Computer Science Department of Carnegie Mellon University as a visiting scientist from Kyoto University in Japan. My plan for research during the stay was to develop a model based object recognition ....

[Article contains additional citation context not shown here]

T. Kanade, Recovery of the three-dimensional shape of an object from a single view, Arttf. Intell. 17 ( 1981 ) 409-460. {3] T. Kanade, Computer vision as a physical science, in: R. Rashid, ed., Carnegie Mellon Computer Science: A 25-Year Commemorative (Addison-Wesley, Reading, MA, 1991) 345-369.

....For example, Schmid and Zisserman used intensity neighborhoods of the line as grey level information [11] Still, it cannot distinguish closely spaced line segments. Because it has difficulty in setting the correct size of a rectangular area to cover the neighborhoods of only a single line. Kanade[6] showed that an edge profile, a curve showing how the intensity changes across an edge, is a good way to find a similar line segment. Setting up a suitable length for extracting an edge profile, however, is difficult when lines are very closely spaced. We propose another attribute, the line ....

Kanade, T., 1981. Recovery of the ThreeDimensional Shape of an Object from a Single View. Artificial Intelligence, 17, pp. 409-460.

No context found.

Kanade, T. 1981. Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence, 17:409--460.

No context found.

T. Kanade, Recovery of the Three-Dimensional Shape of an Object from a Single View, Artificial Intelligence 17, 409--460, 1981.

No context found.

T.Kanade, Recovery of the Three-Dimensional Shape of an Object from a Single View, Artificial Intelligence 17, 409--460, 1981.

No context found.

T. KANADE, 1981. Recovery of the Three-Dimensional Shape of an Object from a Single View, Artificial Intelligence 17, 409-460.

No context found.

T. Kanade. Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence, 17:409--460, 1981.

No context found.

T. Kanade. Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence, 17:409--460, 1981.

No context found.

T. Kanade. Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence, 17:409--460, 1981.

No context found.

T.Kanade. Recovery of the three-dimensional shape of an object from a single view. Artif. Intell, 17:409--460, 1981.

No context found.

T. Kanade, Recovery of the Three-Dimensional Shape of an Object from a Single View, Artificial Intelligence 17, 409--460, 1981.

No context found.

T. Kanade, "Recovery of the three-dimensional shape of an object from a single view," Artificial Intelligence, vol. 17, pp. 409--460, 1981.

No context found.

T.Kanade. Recovery of the Three-Dimensional Shape of an Object from a Single View. Artificial Intelligence 17, 409--460, 1981.

No context found.

T.Kanade, Recovery of the Three-Dimensional Shape of an Object from a Single View, Artificial Intelligence 17, 409--460, 1981.

No context found.

T Kanade. Recovery of the three-dimensional shape of an object from a single view. Artificial Intelligence, 17:409--460, 1981.

No context found.

Kanade T. "Recovery of the three dimensional shape of an object from a single view," Artificial Intelligence 17 (1981), 409-460.

No context found.

T. Kanade, "Recovery of the Three-Dimensional Shape of an Object from a Single View," Artificial Intelligence, Vol. 17, pp. 409-460, 1981.

No context found.

T. Kanade, "Recovery of the three-dimensional shape of an object from a single view", Artificial Intelligence, 17 (1981) p. 409-460.

No context found.

Kanade, T. 1981. Recovery of Three Dimensional Shape of an Object from a Single view. Artificial Intelligence, 17, 409-460.

First 50 documents  Next 50

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