The Curvature Primal Sketch

Asada, H. Harry; Brady, Michael

doi:10.1109/tpami.1986.4767747

Cited by 598 publications

(235 citation statements)

References 19 publications

Supporting

Mentioning

230

Contrasting

Unclassified

Order By: Relevance

“…A common method to segment and describe contours for shape representation is the curvature scale space (Asada and Brady, 1986). Asada and Brady low-pass filter contours with Gaussian functions at increasing scales to detect points of highest curvature at which the contours are broken up and then classified.…”

Section: Previous Contour Approachesmentioning

confidence: 99%

“…Consequently, it may not be ideal to "completely" or "perfectly" partition contours as it has been pursued in shape recognition studies (e.g. (Asada and Brady, 1986;Mokhtarian and Mackworth, 1986)). Rather one should attempt to include this irregularity (or 'wiggliness') as part of the partitioning process.…”

Section: Differing Aspectsmentioning

confidence: 99%

“…b) Late contour classification: Many studies perform a contour classification after curve partitioning: e.g. Asada and Brady classify contours into corners and smooth join, which in turn are used to create compound changes (crank, end, bump or dent) (Asada and Brady, 1986); Bengtsson and Eklundh classify them into straight arcs, curved arcs with sign of curvature, corners (tangent discontinuities) and points delimiting arcs, especially inflexion points (Bengtsson and Eklundh, 1991). As already mentioned above, this classification may prohibit the development of representations capable of buffering variability.…”

Section: Differing Aspectsmentioning

confidence: 99%

“…The common method to do this is the symmetric-axis transform (SAT) (Blum, 1973), which has already been used in many ways. Asada and Brady generate smoothed local symmetries (Asada and Brady, 1986), by taking the pair-wise distances between pixels of parallel contours, a straightforward method to generate the symmetric axis (sym-axis). In other studies, the SAT model aims at evolving sym-axes which represent the shape with high accuracy, applied for instance to the analysis and/or representation of medical image structures, e.g.…”

Section: Region Descriptionmentioning

confidence: 99%

See 3 more Smart Citations

An Approach to the Parameterization of Structure for Fast Categorization

Rasche

2009

Int J Comput Vis

View full text Add to dashboard Cite

A decomposition is described, which parameterizes the geometry and appearance of contours and regions of gray-scale images with the goal of fast categorization. To express the contour geometry, a contour is transformed into a local/global space, from which parameters are derived classifying its global geometry (arc, inflexion or alternating) and describing its local aspects (degree of curvature, edginess, symmetry). Regions are parameterized based on their symmetric axes, which are evolved with a wave-propagation process enabling to generate the distance map for fragmented contour images. The methodology is evaluated on three image sets, the Caltech 101 set and two sets drawn from the Corel collection. The performance nearly reaches the one of other categorization systems for unsupervised learning.

show abstract

Section: Previous Contour Approachesmentioning

confidence: 99%

Section: Differing Aspectsmentioning

confidence: 99%

Section: Differing Aspectsmentioning

confidence: 99%

Section: Region Descriptionmentioning

confidence: 99%

See 2 more Smart Citations

An Approach to the Parameterization of Structure for Fast Categorization

Rasche

2009

Int J Comput Vis

View full text Add to dashboard Cite

show abstract

“…We retrieve them by first calculating a curvature measure for each point and then finding the local maxima. Although there exist sophisticated methods to calculate curvature (see for instance [2]), we have adopted a very simple approach for the sake of efficiency. For each point we consider two more points, one before and one after the current point.…”

Section: Splitting the Linesmentioning

confidence: 99%

A Constructive Feature Detection Approach for Robotic Vision

Hundelshausen

Schreiber

Rojas

2005

RoboCup 2004: Robot Soccer World Cup VIII

View full text Add to dashboard Cite

Abstract. We describe a new method for detecting features on a marked RoboCup field. We implemented the framework for robots with omnidirectional vision, but the method can be easily adapted to other systems. The focus is on the recognition of the center circle and four different corners occurring in the penalty area. Our constructive approach differs from previous methods, in that we aim to detect a whole palette of different features, hierarchically ordered and possibly containing each other. Highlevel features, such as the center circle or the corners, are constructed from low-level features such as arcs and lines. The feature detection process starts with low-level features and iteratively constructs higher features. In RoboCup the method is valuable for robot self-localization; in other fields of application the method is useful for object recognition using shape information.

show abstract

Feature Extraction

Medioni

1999

Wiley Encyclopedia of Electrical and Electronics Engineering

View full text Add to dashboard Cite

show abstract

The Curvature Primal Sketch

Cited by 598 publications

References 19 publications

An Approach to the Parameterization of Structure for Fast Categorization

An Approach to the Parameterization of Structure for Fast Categorization

A Constructive Feature Detection Approach for Robotic Vision

Feature Extraction

Contact Info

Product

Resources

About