#### DMCA

## A Survey of Shape Analysis Techniques (1998)

### Cached

### Download Links

- [web.njit.edu]
- [petra.zesoi.fer.hr]
- CiteULike
- DBLP

### Other Repositories/Bibliography

Venue: | Pattern Recognition |

Citations: | 258 - 2 self |

### Citations

2892 |
The Fractal Geometry of Nature
- Mandelbrot
- 1982
(Show Context)
Citation Context ...evel of control all levels bellow any given level are also included in information processing. Pentland [99, 100] investigated methods for representation of natural forms by means of fractal geometry =-=[101, 102, 103]-=-. He argued that fractal functions are appropriate for natural shape representation because many physical processes produce fractal surface shapes. This is due to the fact that natural forms repeat wh... |

1706 |
Computer Vision
- Horn
- 1987
(Show Context)
Citation Context ...scriptor vectors using a metric. The problem of the shape analysis has been pursued by many authors thus resulting in a great amount of research. A number of review papers [1, 2, 3], as well as books =-=[4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16]-=- have been written on the subject of shape analysis. 1.1 Classi cations Shape analysis methods can be classi ed according to many di erent criteria. Pavlidis [1] has proposed the following classi cati... |

1527 |
The Organization of Behavior
- Hebb
- 1949
(Show Context)
Citation Context ...operties in di erent directions. Another approach to the theory of visual form is found in Hebb's work. Hebb presented a neuropsychological theory of behavior in his book The Organization of Behavior =-=[59]-=-. In his theory, Hebb emphasized the role of neural structures in the mechanism of visual perception. His work in uenced a number of researchers in the eld of arti cial neural networks. As opposed to ... |

1270 | Theory of edge detection
- Marr, Hildreth
- 1980
(Show Context)
Citation Context ...ave no corresponding stimulus in the real world). A tutorial on visual cognition with an emphasis on shape recognition was written by Pinker [61]. 2.2 Modern Theories of Visual Perception Marr et al. =-=[62, 63, 43, 64, 65, 66]-=- made signi cant contributions to the study of the human visual perception system. In Marr's paradigm [67], the focus of research is shifted from applications to topics corresponding to modules of the... |

1242 |
Visual pattern recognition by moment invariants
- Hu
- 1962
(Show Context)
Citation Context ...7, 28, 29]. Examples of global methods include the medial axis (also called symmetric axis) transform (MAT) proposed by Blum and described in [30, 31, 32, 17, 33, 34, 20, 22], moment based approaches =-=[35, 36, 37, 38]-=-, and methods of shape decomposition into other primitive shapes [39, 40, 41, 42]. Another classi cation of shape analysis algorithms can be made on the basis of whether the result of the analysis is ... |

1023 | Computer Vision
- Ballard, Brown
- 1982
(Show Context)
Citation Context ...scriptor vectors using a metric. The problem of the shape analysis has been pursued by many authors thus resulting in a great amount of research. A number of review papers [1, 2, 3], as well as books =-=[4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16]-=- have been written on the subject of shape analysis. 1.1 Classi cations Shape analysis methods can be classi ed according to many di erent criteria. Pavlidis [1] has proposed the following classi cati... |

1001 |
Fractal Everywhere
- Barnsley
- 1988
(Show Context)
Citation Context ...evel of control all levels bellow any given level are also included in information processing. Pentland [99, 100] investigated methods for representation of natural forms by means of fractal geometry =-=[101, 102, 103]-=-. He argued that fractal functions are appropriate for natural shape representation because many physical processes produce fractal surface shapes. This is due to the fact that natural forms repeat wh... |

992 |
An Introduction to Signal Detection and Estimation
- Poor
- 1998
(Show Context)
Citation Context ...quation rt = + mX j=1 jrt,j + p !t where !i are independent random noise sources. The parameters f ; 1;:::; m; g are unknown and need to be estimated. The maximum likelihood (ML) parameter estimation =-=[113, 114, 115]-=- was used. The ML estimated parameters i are translation, rotation, and scale invariant. Note that the rotation invariance holds only for angles that are multiples of 2 =N . Parameters and are not sca... |

840 |
The structure of images
- Koenderink
- 1984
(Show Context)
Citation Context ... crossings of the Laplacian of the two-dimensional Gaussian distribution function. In this approach, zeros of Laplacian indicate the in ection point in the edge to detect edge positions. 8sKoenderink =-=[73]-=- and Koenderink and van Doorn [74] have studied the psychological aspects of visual perception and proposed several interesting paradigms. Conventional approaches to shape are often static in the sens... |

827 |
Pattern Recognition. A Statistical Approach
- Devijver, Kittler
- 1992
(Show Context)
Citation Context ...groups the members of one pattern class closer in the new coordinate system. The third modi cation included the use of hyper-planes to divide the pattern space. The least mean squared error procedure =-=[118]-=- yields the optimal hyper-plane parameters. The experimental result showed that the normalized AR model parameters useful shape descriptors. A modi cation of the Dubois and Glanz method included stati... |

633 |
A Transformation for Extracting New Descriptors of Shape. Models for the Perception
- Blum
- 1967
(Show Context)
Citation Context ... and various Fourier transforms of the boundary [25, 26, 27, 28, 29]. Examples of global methods include the medial axis (also called symmetric axis) transform (MAT) proposed by Blum and described in =-=[30, 31, 32, 17, 33, 34, 20, 22]-=-, moment based approaches [35, 36, 37, 38], and methods of shape decomposition into other primitive shapes [39, 40, 41, 42]. Another classi cation of shape analysis algorithms can be made on the basis... |

624 | Some informational aspects of visual perception
- Attneave
- 1954
(Show Context)
Citation Context ...di erential equation that describes the evolution of the shape through multiple resolutions. Many authors agree on the signi cance of high curvature points for visual perception. Attneave and Arnoult =-=[76, 77]-=- performed psychological experiments to investigate the signi cance of corners for perception. In the famous Attneave's cat experiment a drawing of a cat was used to locate points of high curvature wh... |

618 |
Representation and recognition of the spatial organization of three-dimensional shapes
- Marr, Nishihara
- 1978
(Show Context)
Citation Context ...ll do not exist for shape description methods. Several authors have proposed evaluation criteria in the form of lists of qualities that a good shape description method should have. Marr and Nishihara =-=[43]-=- proposed a set of criteria for the evaluation of shape description methods: accessibility, scope and uniqueness, and stability and sensitivity. Accessibility describes how easy (or di cult) it is to ... |

605 |
The Perception of the Visual World
- Gibson
- 1950
(Show Context)
Citation Context ...s causing neurons in the group to mutually excite each other. Hebb's theory was mostly qualitative and not computational, thus presenting a disadvantage for practical engineering applications. Gibson =-=[60]-=- developed another comprehensive theory of visual perception. The rst principle of his theory is that space is not a geometric or abstract entity, but a real visual one characterized by the forms that... |

524 |
Perceptual Organization and Visual Recognition
- Lowe
- 1985
(Show Context)
Citation Context ...An exhaustive survey of human visual perception research is beyond the scope of this paper. Some introductory and more advanced books and articles dealing with visual perception and cognition include =-=[4, 46, 47,48,49,50,51,52]-=-. In this section, a brief overview of visual perception research related to shape description is presented. 2.1 Classical Theories of Visual Perception Several schools of psychology have endeavored t... |

480 | Three-dimensional object recognition from single two-dimensional images
- Lowe
(Show Context)
Citation Context ...oped for engineering purposes and not necessarily to study perception. Fractal representations produce objects which correspond much better to the human model of visual perception and cognition. Lowe =-=[106]-=- proposed a computer vision system that can recognize three-dimensional objects from unknown viewpoints and single two-dimensional images. The procedure is non-typical and uses three mechanisms of per... |

432 |
T.: A computational theory of human stereo vision
- Marr, Poggio
- 1979
(Show Context)
Citation Context ...ave no corresponding stimulus in the real world). A tutorial on visual cognition with an emphasis on shape recognition was written by Pinker [61]. 2.2 Modern Theories of Visual Perception Marr et al. =-=[62, 63, 43, 64, 65, 66]-=- made signi cant contributions to the study of the human visual perception system. In Marr's paradigm [67], the focus of research is shifted from applications to topics corresponding to modules of the... |

334 | Fourier descriptors for plane closed curves
- Zahn, Roskies
- 1972
(Show Context)
Citation Context ...nal) and global (or internal), respectively. Examples of the former class are algorithms which parse the shape boundary [17, 18, 19, 20, 21, 22, 23, 24] and various Fourier transforms of the boundary =-=[25, 26, 27, 28, 29]-=-. Examples of global methods include the medial axis (also called symmetric axis) transform (MAT) proposed by Blum and described in [30, 31, 32, 17, 33, 34, 20, 22], moment based approaches [35, 36, 3... |

318 |
Image analysis via the general theory of moments
- Teague
- 1979
(Show Context)
Citation Context ...7, 28, 29]. Examples of global methods include the medial axis (also called symmetric axis) transform (MAT) proposed by Blum and described in [30, 31, 32, 17, 33, 34, 20, 22], moment based approaches =-=[35, 36, 37, 38]-=-, and methods of shape decomposition into other primitive shapes [39, 40, 41, 42]. Another classi cation of shape analysis algorithms can be made on the basis of whether the result of the analysis is ... |

269 |
Shape description using weighted symmetric axis features
- Blum, Nagel
- 1978
(Show Context)
Citation Context ... and various Fourier transforms of the boundary [25, 26, 27, 28, 29]. Examples of global methods include the medial axis (also called symmetric axis) transform (MAT) proposed by Blum and described in =-=[30, 31, 32, 17, 33, 34, 20, 22]-=-, moment based approaches [35, 36, 37, 38], and methods of shape decomposition into other primitive shapes [39, 40, 41, 42]. Another classi cation of shape analysis algorithms can be made on the basis... |

267 |
Pattern Classi� cation and Scene Analysis
- Duda, Hart
- 1973
(Show Context)
Citation Context ...scriptor vectors using a metric. The problem of the shape analysis has been pursued by many authors thus resulting in a great amount of research. A number of review papers [1, 2, 3], as well as books =-=[4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16]-=- have been written on the subject of shape analysis. 1.1 Classi cations Shape analysis methods can be classi ed according to many di erent criteria. Pavlidis [1] has proposed the following classi cati... |

256 |
Human image understanding: recent research and a theory
- Biederman
- 1985
(Show Context)
Citation Context ...An exhaustive survey of human visual perception research is beyond the scope of this paper. Some introductory and more advanced books and articles dealing with visual perception and cognition include =-=[4, 46, 47,48,49,50,51,52]-=-. In this section, a brief overview of visual perception research related to shape description is presented. 2.1 Classical Theories of Visual Perception Several schools of psychology have endeavored t... |

251 |
Shape discrimination using Fourier Descriptors
- Persoon, Fu
- 1977
(Show Context)
Citation Context ...nal) and global (or internal), respectively. Examples of the former class are algorithms which parse the shape boundary [17, 18, 19, 20, 21, 22, 23, 24] and various Fourier transforms of the boundary =-=[25, 26, 27, 28, 29]-=-. Examples of global methods include the medial axis (also called symmetric axis) transform (MAT) proposed by Blum and described in [30, 31, 32, 17, 33, 34, 20, 22], moment based approaches [35, 36, 3... |

243 |
Fractal-based description of natural scenes
- Pentland
- 1984
(Show Context)
Citation Context ...erns the functioning of the human perceptual system. It is a hierarchical system where at each level of control all levels bellow any given level are also included in information processing. Pentland =-=[99, 100]-=- investigated methods for representation of natural forms by means of fractal geometry [101, 102, 103]. He argued that fractal functions are appropriate for natural shape representation because many p... |

237 |
Visual Perception
- Cornsweet
- 1970
(Show Context)
Citation Context ...An exhaustive survey of human visual perception research is beyond the scope of this paper. Some introductory and more advanced books and articles dealing with visual perception and cognition include =-=[4, 46, 47,48,49,50,51,52]-=-. In this section, a brief overview of visual perception research related to shape description is presented. 2.1 Classical Theories of Visual Perception Several schools of psychology have endeavored t... |

233 |
An Efficiently Computable Metric for Comparing Polygonal Shapes
- Arkin
- 1991
(Show Context)
Citation Context ...arc length function. The tangent angle at some point is measured relative to the tangent angle at the initial point. The function is also called the turning function and has been used by Arkin et al. =-=[108]-=- for comparing polygonal shapes. 11sO A4 A3 A2 A1 A0 Figure 1: The centroid-to-boundary distance approach. Granlund [26], Richards and Hemami [27], and Persoon and Fu [28] used a complex function of t... |

189 |
Perceptual organization and the representation of natural form
- Pentland
- 1986
(Show Context)
Citation Context ...erns the functioning of the human perceptual system. It is a hierarchical system where at each level of control all levels bellow any given level are also included in information processing. Pentland =-=[99, 100]-=- investigated methods for representation of natural forms by means of fractal geometry [101, 102, 103]. He argued that fractal functions are appropriate for natural shape representation because many p... |

173 |
Solid Shape
- Koenderink
- 1990
(Show Context)
Citation Context ... two-dimensional Gaussian distribution function. In this approach, zeros of Laplacian indicate the in ection point in the edge to detect edge positions. 8sKoenderink [73] and Koenderink and van Doorn =-=[74]-=- have studied the psychological aspects of visual perception and proposed several interesting paradigms. Conventional approaches to shape are often static in the sense that they treat all shape detail... |

150 |
Laws of organization in perceptual forms
- Wertheimer
- 1938
(Show Context)
Citation Context ...f selected topics. The Gestalt school of psychology [54] has played a revolutionary role with its novel approach to visual form. A detailed exposition of the subject of Gestalt theory can be found in =-=[55, 56, 57, 58]. The-=- Gestalt theory is a non-computational theory of visual form, and thus a disadvantage for practical engineering applications. However, according to Zusne "it is still the only theory to deal with... |

150 |
Early processing of visual information
- Marr
- 1976
(Show Context)
Citation Context ...ave no corresponding stimulus in the real world). A tutorial on visual cognition with an emphasis on shape recognition was written by Pinker [61]. 2.2 Modern Theories of Visual Perception Marr et al. =-=[62, 63, 43, 64, 65, 66]-=- made signi cant contributions to the study of the human visual perception system. In Marr's paradigm [67], the focus of research is shifted from applications to topics corresponding to modules of the... |

149 |
Directional selectivity and its use in early visual processing
- Marr, Ullman
- 1981
(Show Context)
Citation Context |

147 | A survey of moment-based techniques for unoccluded object representation and recognition
- Prokop, Reeves
- 1992
(Show Context)
Citation Context ...7, 28, 29]. Examples of global methods include the medial axis (also called symmetric axis) transform (MAT) proposed by Blum and described in [30, 31, 32, 17, 33, 34, 20, 22], moment based approaches =-=[35, 36, 37, 38]-=-, and methods of shape decomposition into other primitive shapes [39, 40, 41, 42]. Another classi cation of shape analysis algorithms can be made on the basis of whether the result of the analysis is ... |

140 |
On the detection of dominant points on digital curves
- Teh, Chin
- 1989
(Show Context)
Citation Context ...at could be reliably recognized in the drawing. It has been suggested that such points have high information content. Attneave's work has initiated further research on the topic of curve partitioning =-=[78, 79, 80, 81, 82, 83, 84, 85]-=-. To approximate curves by straight lines, high curvature points are the best place to break the lines, thereby the resulting image retaining in the maximal amount of information necessary for success... |

136 | Smoothed local symmetries and their implementation
- Brady, Asada
- 1984
(Show Context)
Citation Context ...f representation. Brady illustrated these criteria on the Generalized Cylinder representation for three-dimensional objects and the Smoothed Local Symmetries representation for two-dimensional shapes =-=[45]-=-. Both of the methods mentioned above de ne desired properties in terms of conceptual qualities that cannot be numerically expressed. This limitation renders an exact numerical comparison of shape des... |

110 |
Segmentation of plane curves
- Pavlidis, Horovitz
- 1974
(Show Context)
Citation Context ...at could be reliably recognized in the drawing. It has been suggested that such points have high information content. Attneave's work has initiated further research on the topic of curve partitioning =-=[78, 79, 80, 81, 82, 83, 84, 85]-=-. To approximate curves by straight lines, high curvature points are the best place to break the lines, thereby the resulting image retaining in the maximal amount of information necessary for success... |

99 |
Fourier preprocessing for hand print character recognition
- Granlund
- 1972
(Show Context)
Citation Context ...nal) and global (or internal), respectively. Examples of the former class are algorithms which parse the shape boundary [17, 18, 19, 20, 21, 22, 23, 24] and various Fourier transforms of the boundary =-=[25, 26, 27, 28, 29]-=-. Examples of global methods include the medial axis (also called symmetric axis) transform (MAT) proposed by Blum and described in [30, 31, 32, 17, 33, 34, 20, 22], moment based approaches [35, 36, 3... |

89 | N.: Surfaces from stereo: integrating feature matching, disparity estimation and contour detection
- Hoff, Ahuja
- 1989
(Show Context)
Citation Context ...art of the total research in computer vision [3]. Papers dealing with shape from x techniques include: shape from shading [68], shape from contour [69, 70], shape from texture [60], shape from stereo =-=[71]-=-, and shape from fractal geometry [72]. In [63] Marr developed a primal sketch paradigm for early processing of visual information. In his method, a set of masks is used to measure discontinuities in ... |

88 |
Pattern recognition with moment invariants: a comparative study and new results, Pattern Recognition 24
- Belkasim, Shridhar, et al.
- 1991
(Show Context)
Citation Context |

86 |
Angle detection on digital curves
- Rosenfeld, Johnston
- 1973
(Show Context)
Citation Context ...at could be reliably recognized in the drawing. It has been suggested that such points have high information content. Attneave's work has initiated further research on the topic of curve partitioning =-=[78, 79, 80, 81, 82, 83, 84, 85]-=-. To approximate curves by straight lines, high curvature points are the best place to break the lines, thereby the resulting image retaining in the maximal amount of information necessary for success... |

83 |
Visual cognition: An introduction
- Pinker
- 1984
(Show Context)
Citation Context ...mbols but do not correspond to real objects (because they have no corresponding stimulus in the real world). A tutorial on visual cognition with an emphasis on shape recognition was written by Pinker =-=[61]-=-. 2.2 Modern Theories of Visual Perception Marr et al. [62, 63, 43, 64, 65, 66] made signi cant contributions to the study of the human visual perception system. In Marr's paradigm [67], the focus of ... |

74 |
A Theory for Cerebral Neocortex
- Marr
- 1970
(Show Context)
Citation Context |

69 |
An efficient threedimensional aircraft recognition algorithm using normalized Fourier descriptors
- WALLACE, WINTZ
- 1980
(Show Context)
Citation Context |

67 |
Stochastic Processes in Engineering Systems
- Wong, Hajek
- 1985
(Show Context)
Citation Context ...quation rt = + mX j=1 jrt,j + p !t where !i are independent random noise sources. The parameters f ; 1;:::; m; g are unknown and need to be estimated. The maximum likelihood (ML) parameter estimation =-=[113, 114, 115]-=- was used. The ML estimated parameters i are translation, rotation, and scale invariant. Note that the rotation invariance holds only for angles that are multiples of 2 =N . Parameters and are not sca... |

58 |
A review of algorithms for shape analysis
- Pavlidis
- 1978
(Show Context)
Citation Context ...es by comparing the shape descriptor vectors using a metric. The problem of the shape analysis has been pursued by many authors thus resulting in a great amount of research. A number of review papers =-=[1, 2, 3]-=-, as well as books [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16] have been written on the subject of shape analysis. 1.1 Classi cations Shape analysis methods can be classi ed according to many di er... |

55 | The quantitative study of shape and pattern perception
- Attneave, Arnoult
- 1956
(Show Context)
Citation Context ...di erential equation that describes the evolution of the shape through multiple resolutions. Many authors agree on the signi cance of high curvature points for visual perception. Attneave and Arnoult =-=[76, 77]-=- performed psychological experiments to investigate the signi cance of corners for perception. In the famous Attneave's cat experiment a drawing of a cat was used to locate points of high curvature wh... |

54 |
Criteria for representations of shape
- Brady
- 1983
(Show Context)
Citation Context ...niqueness describes whether a one-to-one mapping exists between shapes and shape descriptors. Stability and sensitivity describe how sensitive a shape descriptor is to "small" changes in sha=-=pe. Brady [44]-=- proposed another set of criteria for the representation of shape. Rich local support. This criterion requires that a representation is information preserving (rich) and can be computed locally. Local... |

52 |
A source book of Gestalt psychology
- Ellis
- 1938
(Show Context)
Citation Context ...f selected topics. The Gestalt school of psychology [54] has played a revolutionary role with its novel approach to visual form. A detailed exposition of the subject of Gestalt theory can be found in =-=[55, 56, 57, 58]. The-=- Gestalt theory is a non-computational theory of visual form, and thus a disadvantage for practical engineering applications. However, according to Zusne "it is still the only theory to deal with... |

51 |
Algorithm for Shape Analysis of Contours and Waveforms
- Pavlidis
- 1978
(Show Context)
Citation Context ...es by comparing the shape descriptor vectors using a metric. The problem of the shape analysis has been pursued by many authors thus resulting in a great amount of research. A number of review papers =-=[1, 2, 3]-=-, as well as books [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16] have been written on the subject of shape analysis. 1.1 Classi cations Shape analysis methods can be classi ed according to many di er... |

50 | Codon constraints on closed 2D shapes
- Richards, Hoffman
- 1985
(Show Context)
Citation Context ...pe approximations. Stokely and Shang [93] investigated methods for measurement of the curvature of 3-D surfaces that evolve in many applications (e.g. tomographic medical images). Ho man and Richards =-=[94, 95]-=- argue that when the visual system decomposes objects it does so at points of high negative curvature. This agrees with the principle of transversality [96] found in nature. The principle 9sof transve... |

45 |
Computational-geometric methods for polygonal approximations of a curve
- Imai, Iri
- 1986
(Show Context)
Citation Context ...nition. For the purpose of shape description, corners are used as points of high curvature and the shape can be approximated by a polygon. A variety of algorithms for polygonal approximation of shape =-=[86, 1, 87, 88, 89, 90, 91]-=- have been developed. Davis [92] combined the use of high curvature points and line segment approximations for polygonal shape approximations. Stokely and Shang [93] investigated methods for measureme... |

42 |
Stochastic models for closed boundary analysis: Representation and reconstruction
- Kashyap, Chellappa
- 1981
(Show Context)
Citation Context ...The two resulting classes of algorithms are known as boundary (also called external) and global (or internal), respectively. Examples of the former class are algorithms which parse the shape boundary =-=[17, 18, 19, 20, 21, 22, 23, 24]-=- and various Fourier transforms of the boundary [25, 26, 27, 28, 29]. Examples of global methods include the medial axis (also called symmetric axis) transform (MAT) proposed by Blum and described in ... |

39 | Symmetry-curvature duality
- Leyton
- 1987
(Show Context)
Citation Context ... 9sof transversality contends that when two arbitrarily shaped convex objects interpenetrate each other, the meeting point is a boundary point of concave discontinuity of their tangent planes. Leyton =-=[97]-=- demonstrated the Symmetry-Curvature theorem which claims that any curve section that has only one curvature extremum has one and only one symmetric axis which terminates at the extremum itself. (For ... |

36 |
Robust contour decomposition using constant curvature criterion
- Wuescher, Boyer
- 1991
(Show Context)
Citation Context |

36 |
An Autoregressive Model Approach to TwoDimensional Shape Classification
- Dubois, Glanz
- 1986
(Show Context)
Citation Context ...riptor. Kashyap and Chellappa further investigated coding and reconstruction schemes and showed that stochastic methods could be used for the description of closed boundaries. 14 (2)sDubois and Glanz =-=[116]-=- used the same autoregressive model as in [19] but investigated three additional methods for improving pattern classi cation (shape matching) performance. The classi cation was then performed by compu... |

35 |
Decomposition of three-dimensional objects into spheres
- O’Rourke, Badler
- 1979
(Show Context)
Citation Context ...tric axis) transform (MAT) proposed by Blum and described in [30, 31, 32, 17, 33, 34, 20, 22], moment based approaches [35, 36, 37, 38], and methods of shape decomposition into other primitive shapes =-=[39, 40, 41, 42]-=-. Another classi cation of shape analysis algorithms can be made on the basis of whether the result of the analysis is numeric or non-numeric. For example, the MAT produces another image (containing a... |

35 |
Perceptual Organization and Curve Partitioning
- Fischler, Bolles
- 1986
(Show Context)
Citation Context |

34 |
Segmenting handwritten signatures at their perceptually important points
- Brault, Plamondon
- 1993
(Show Context)
Citation Context |

34 |
Inferring causal history from shape
- Leyton
- 1989
(Show Context)
Citation Context ...xtremum itself. (For more information on symmetric axis work see Section 6.1.) This is an important result because it establishes the connection between two important notions in visual perception. In =-=[98]-=-, Leyton developed a theory which claims that all shapes are basically circles which changed form as a result of various deformations caused by external forces like pushing, pulling, stretching, etc. ... |

32 |
An Analysis Technique for Biological Shape
- Young, Walker, et al.
- 1974
(Show Context)
Citation Context ...The two resulting classes of algorithms are known as boundary (also called external) and global (or internal), respectively. Examples of the former class are algorithms which parse the shape boundary =-=[17, 18, 19, 20, 21, 22, 23, 24]-=- and various Fourier transforms of the boundary [25, 26, 27, 28, 29]. Examples of global methods include the medial axis (also called symmetric axis) transform (MAT) proposed by Blum and described in ... |

31 |
Biological shape and visual science (part 1
- Blum
- 1973
(Show Context)
Citation Context ... and various Fourier transforms of the boundary [25, 26, 27, 28, 29]. Examples of global methods include the medial axis (also called symmetric axis) transform (MAT) proposed by Blum and described in =-=[30, 31, 32, 17, 33, 34, 20, 22]-=-, moment based approaches [35, 36, 37, 38], and methods of shape decomposition into other primitive shapes [39, 40, 41, 42]. Another classi cation of shape analysis algorithms can be made on the basis... |

31 |
Understanding shape: Angles and sides
- Davis
- 1977
(Show Context)
Citation Context ...re used as points of high curvature and the shape can be approximated by a polygon. A variety of algorithms for polygonal approximation of shape [86, 1, 87, 88, 89, 90, 91] have been developed. Davis =-=[92]-=- combined the use of high curvature points and line segment approximations for polygonal shape approximations. Stokely and Shang [93] investigated methods for measurement of the curvature of 3-D surfa... |

27 |
Polygonal approximation by Newton’s method
- Pavlidis
- 1977
(Show Context)
Citation Context ...nition. For the purpose of shape description, corners are used as points of high curvature and the shape can be approximated by a polygon. A variety of algorithms for polygonal approximation of shape =-=[86, 1, 87, 88, 89, 90, 91]-=- have been developed. Davis [92] combined the use of high curvature points and line segment approximations for polygonal shape approximations. Stokely and Shang [93] investigated methods for measureme... |

27 |
Dierential Topology
- Guillemin, Pollack
- 1974
(Show Context)
Citation Context ...ic medical images). Ho man and Richards [94, 95] argue that when the visual system decomposes objects it does so at points of high negative curvature. This agrees with the principle of transversality =-=[96]-=- found in nature. The principle 9sof transversality contends that when two arbitrarily shaped convex objects interpenetrate each other, the meeting point is a boundary point of concave discontinuity o... |

26 |
A shape recognition scheme based on relative distances of feature points from the centroid
- Chang, Hwang, et al.
- 1991
(Show Context)
Citation Context ... is approximated with a polygon so that all sides have the same length. The values of the 1-D function are equal to the angle between the polygon sides and a reference line. Chang, Hwang, and Buehrer =-=[109]-=- constructed the distance function from the centroid to the feature points. In their method, the feature points are the points of high curvature. There are two approaches for detecting high curvature ... |

24 |
Locating perceptually salient points on planar curves
- Fischler, Wolf
- 1994
(Show Context)
Citation Context |

23 |
Visual shape computation
- Aloimonos
- 1988
(Show Context)
Citation Context ...es by comparing the shape descriptor vectors using a metric. The problem of the shape analysis has been pursued by many authors thus resulting in a great amount of research. A number of review papers =-=[1, 2, 3]-=-, as well as books [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16] have been written on the subject of shape analysis. 1.1 Classi cations Shape analysis methods can be classi ed according to many di er... |

23 |
On the geometric interpretation of image contours
- Horaud, Brady
- 1987
(Show Context)
Citation Context ...shape from x research which represents an important part of the total research in computer vision [3]. Papers dealing with shape from x techniques include: shape from shading [68], shape from contour =-=[69, 70]-=-, shape from texture [60], shape from stereo [71], and shape from fractal geometry [72]. In [63] Marr developed a primal sketch paradigm for early processing of visual information. In his method, a se... |

23 |
Decomposition of Convex Polygonal Morphological Structuring Elements into Neighborhood Subsets
- Xu
- 1991
(Show Context)
Citation Context ...nition. For the purpose of shape description, corners are used as points of high curvature and the shape can be approximated by a polygon. A variety of algorithms for polygonal approximation of shape =-=[86, 1, 87, 88, 89, 90, 91]-=- have been developed. Davis [92] combined the use of high curvature points and line segment approximations for polygonal shape approximations. Stokely and Shang [93] investigated methods for measureme... |

23 |
On Growth and Form
- Thomson
- 2000
(Show Context)
Citation Context ...tion because many physical processes produce fractal surface shapes. This is due to the fact that natural forms repeat whenever possible and non-animal objects have a limited number of possible forms =-=[104, 105]-=-. Most existing schemes for shape representation were developed for engineering purposes and not necessarily to study perception. Fractal representations produce objects which correspond much better t... |

20 |
Machine Perception
- Nevatia
- 1982
(Show Context)
Citation Context |

20 |
Visuai Perception of Form
- Zusne
- 1970
(Show Context)
Citation Context |

20 |
Surface parametrization and curvature measurement of arbitrary 3-d objects: five practical methods
- Stokely, Wu
- 1992
(Show Context)
Citation Context ...on of shape [86, 1, 87, 88, 89, 90, 91] have been developed. Davis [92] combined the use of high curvature points and line segment approximations for polygonal shape approximations. Stokely and Shang =-=[93]-=- investigated methods for measurement of the curvature of 3-D surfaces that evolve in many applications (e.g. tomographic medical images). Ho man and Richards [94, 95] argue that when the visual syste... |

19 |
A min max medial axis transformation
- Peleg, Rosenfeld
- 1981
(Show Context)
Citation Context |

19 | Music recognition
- unknown authors
- 1969
(Show Context)
Citation Context ...pe approximations. Stokely and Shang [93] investigated methods for measurement of the curvature of 3-D surfaces that evolve in many applications (e.g. tomographic medical images). Ho man and Richards =-=[94, 95]-=- argue that when the visual system decomposes objects it does so at points of high negative curvature. This agrees with the principle of transversality [96] found in nature. The principle 9sof transve... |

17 |
Discrimination of planar shapes using shape matrices
- Taza, Suen
- 1989
(Show Context)
Citation Context ...The two resulting classes of algorithms are known as boundary (also called external) and global (or internal), respectively. Examples of the former class are algorithms which parse the shape boundary =-=[17, 18, 19, 20, 21, 22, 23, 24]-=- and various Fourier transforms of the boundary [25, 26, 27, 28, 29]. Examples of global methods include the medial axis (also called symmetric axis) transform (MAT) proposed by Blum and described in ... |

16 |
The role of human visual models in image processing
- Granrath
- 1981
(Show Context)
Citation Context |

15 |
Polygonal Shape Recognition using String Matching Techniques
- Maes
- 1994
(Show Context)
Citation Context |

14 |
Shape description by time series
- Kartikeyan, Sarkar
- 1989
(Show Context)
Citation Context |

12 |
Minimal rectangular partitions of digitized blobs. Computer vision, graphics, and image processing
- Ferrari, Sankar, et al.
- 1984
(Show Context)
Citation Context ...tric axis) transform (MAT) proposed by Blum and described in [30, 31, 32, 17, 33, 34, 20, 22], moment based approaches [35, 36, 37, 38], and methods of shape decomposition into other primitive shapes =-=[39, 40, 41, 42]-=-. Another classi cation of shape analysis algorithms can be made on the basis of whether the result of the analysis is numeric or non-numeric. For example, the MAT produces another image (containing a... |

12 |
Invariant Pattern Recognition by Moment Fourier Descriptor
- Wang, Chen, et al.
- 1994
(Show Context)
Citation Context ...ieve rotation invariance. The distances are also divided by the minimal distance to achieve scale invariance. Translation invariance is automatically achieved through the use of centroid. Wang et al. =-=[110]-=- used a sequence of line segment moments as a 1-D function. Line segments are obtained 12sby partitioning the radial line from the center of the mass to the boundary point. Segments are partitioned in... |

10 |
Curve segmentation under partial occlusion
- Katzir, Lindenbaum, et al.
- 1994
(Show Context)
Citation Context |

10 |
A solution of polygon containment, spatial planning, and other related problems using minkowski operations
- Ghosh
- 1990
(Show Context)
Citation Context |

8 |
New polygonal approximation schemes for object shape representation, Pattern Recognition 26
- Wu, Leou
- 1993
(Show Context)
Citation Context |

7 |
Models for the perception of speech and visual form
- Wathen-Dunn
- 1967
(Show Context)
Citation Context |

7 |
Infering shape from contour for curved surfaces
- Ulupmar, Nevatia
- 1990
(Show Context)
Citation Context ...shape from x research which represents an important part of the total research in computer vision [3]. Papers dealing with shape from x techniques include: shape from shading [68], shape from contour =-=[69, 70]-=-, shape from texture [60], shape from stereo [71], and shape from fractal geometry [72]. In [63] Marr developed a primal sketch paradigm for early processing of visual information. In his method, a se... |

7 |
Multiscale Fourier descriptors for classifying semivowels in spectrograms
- Pinkowski
- 1993
(Show Context)
Citation Context ...ized in the frequency domain. Tangent angle versus arc length representations su er from very high noise sensitivity because it is di cult to determine the tangent angle for noisy contours. Pinkowski =-=[111]-=- used Fourier descriptors for the description of shapes appearing in speech spectrograms. This method is used for classifying words containing English semi-vowels. Experiments demonstrated a high reco... |

7 |
Coding of two-tone images
- Huang
- 1952
(Show Context)
Citation Context ...ers and an independent noise sequence. Since the boundary is closed, boundary 1-D function rt is a periodic function. The particular CAR model that was utilized is the same one that was used by Huang =-=[112]-=-. It is a stochastic process de ned by the following m-th order di erence equation rt = + mX j=1 jrt,j + p !t where !i are independent random noise sources. The parameters f ; 1;:::; m; g are unknown ... |

6 |
Shape from fractal geometry
- Chen, Keller, et al.
- 1990
(Show Context)
Citation Context ...vision [3]. Papers dealing with shape from x techniques include: shape from shading [68], shape from contour [69, 70], shape from texture [60], shape from stereo [71], and shape from fractal geometry =-=[72]-=-. In [63] Marr developed a primal sketch paradigm for early processing of visual information. In his method, a set of masks is used to measure discontinuities in rst and second derivatives of the orig... |

5 |
Form discrimination and the invariance of form
- HAKE
- 1966
(Show Context)
Citation Context ...ssical Theories of Visual Perception Several schools of psychology have endeavored to understand and describe the mechanisms of behavior, in general, and the speci c aspect of visual perception. Hake =-=[53]-=- discussed several approaches to the representation of natural forms. Zusne [47] presented an overview of contemporary theories of visual form. An in-depth discussion of visual perception theories in ... |

5 |
Principles of Gestalt Psychology
- ka, K
- 1935
(Show Context)
Citation Context ...f selected topics. The Gestalt school of psychology [54] has played a revolutionary role with its novel approach to visual form. A detailed exposition of the subject of Gestalt theory can be found in =-=[55, 56, 57, 58]. The-=- Gestalt theory is a non-computational theory of visual form, and thus a disadvantage for practical engineering applications. However, according to Zusne "it is still the only theory to deal with... |

5 |
On the Measurement of Curvature in a Quantized Environment
- Bennet, Donald, et al.
- 1975
(Show Context)
Citation Context ... a real or complex 1-D function. In this Section we present several possible 1-D boundary representations of the shape that have been used in literature. Zahn and Roskies [25] and Bennet and McDonald =-=[107]-=- used a tangent angle versus arc length function. The tangent angle at some point is measured relative to the tangent angle at the initial point. The function is also called the turning function and h... |

4 |
Description and descrimination of planar shapes using shape matrices
- GOSHTASBY
- 1985
(Show Context)
Citation Context |

4 |
On sequential shape descriptions
- Jagadish, Bruckstein
- 1992
(Show Context)
Citation Context ...tric axis) transform (MAT) proposed by Blum and described in [30, 31, 32, 17, 33, 34, 20, 22], moment based approaches [35, 36, 37, 38], and methods of shape decomposition into other primitive shapes =-=[39, 40, 41, 42]-=-. Another classi cation of shape analysis algorithms can be made on the basis of whether the result of the analysis is numeric or non-numeric. For example, the MAT produces another image (containing a... |

3 |
Two-dimensional shape representation
- Davis
- 1986
(Show Context)
Citation Context |

3 |
Identi cation of threedimensional objects using Fourier descriptors of the boundary curve
- Richard, Hemami
- 1974
(Show Context)
Citation Context |

3 |
Medial axis transformations
- Mott-Smith
- 1970
(Show Context)
Citation Context |

3 |
Contemporary Theory of Visual Form Perception: III. The Global Theories, chapter 4
- Zusne
- 1970
(Show Context)
Citation Context ...depth discussion of visual perception theories in psychology is outside the scope of this work, therefore we limit ourselves to a brief discussion of selected topics. The Gestalt school of psychology =-=[54]-=- has played a revolutionary role with its novel approach to visual form. A detailed exposition of the subject of Gestalt theory can be found in [55, 56, 57, 58]. The Gestalt theory is a non-computatio... |

3 |
Height and Gradient from Shading, Intl
- Horn
- 1990
(Show Context)
Citation Context ...s point is the so-called shape from x research which represents an important part of the total research in computer vision [3]. Papers dealing with shape from x techniques include: shape from shading =-=[68]-=-, shape from contour [69, 70], shape from texture [60], shape from stereo [71], and shape from fractal geometry [72]. In [63] Marr developed a primal sketch paradigm for early processing of visual inf... |

2 |
Shape Analysis and Recognition, chapter 5
- Nevatia
- 1982
(Show Context)
Citation Context |

2 |
Shape decomposition by collinearity
- Kim, Park, et al.
- 1987
(Show Context)
Citation Context |