Computing dominant points on simple shapes

doi:10.1016/S0020-7373(74)80041-6

International Journal of Man-Machine Studies

Volume 6, Issue 1, January 1974, Pages 1-12

https://doi.org/10.1016/S0020-7373(74)80041-6 Get rights and content

Gestalt psychologists have given many examples to demonstrate that laws of visual organization cause one view of a scene to dominate others. This is also true for simple shapes. A figure can be articulated into many fragments but only a few will be perceptually dominant. The analysis of shape requires that the computations underlying the articulation be made explicit. Possible computations have been investigated on convex blobs, the dominant units being corner-like regions. Each region is considered to have a point which perceptually dominates other points in that region. The problem, then, is how to attribute a set of points to some point so that the set is perceived as a meaningful fragment of the blob.

References (9)

T. Pavlidis
Analysis of set patterns
Pat. Recog
(1968)
B. Rosenberg
The analysis of convex blobs
Computer Graphics and Image Processing
(1972)
M.B. Clowes
Picture syntax
H.G. Furth
Piaget and Knowledge
(1969)

There are more references available in the full text version of this article.

Cited by (10)

Recognition of digital curves scanned from paper drawings using genetic algorithms
2003, Pattern Recognition
After analyzing the existing methods, based on holo-extraction method of information, this paper develops a recognition method of digital curves scanned from paper drawings for subsequent pattern recognition and 3D reconstruction. This method is first to construct the networks of single closed region (SCRs) of black pixels with all the information about both segments and their linking points, to classify all the digital contours represented by SCRs into three types: straight-line segments, circular arcs, and combined lines, and then to decompose the combined lines into least basic sub-lines or segments (straight-line segments or circular arcs) with least fitting errors using genetic algorithms with adaptive probabilities of crossover and mutation and to determine their relationships (intersecting or being tangential to each other). It is verified that the recognition method based on the networks of SCRs and the genetic algorithm is feasible and efficient. This method and its software prototype can be used as a base for further work on subsequent engineering drawing understanding and 3D reconstruction.
Reduced resolution and scale space for dominant feature detection in contours
2001, Pattern Recognition
Analysis of contours or curves at various scales or levels of smoothing (scale space) is an important tool in curve segmentation and feature detection. In this paper we propose the application of the hierarchical discrete correlation algorithm for efficiently calculating and creating a scale space of curves. In conjunction with this procedure we also investigate the use of the reduced resolution or Gaussian pyramid representation of the set of smoothed curves as the basis for initially localizing and detecting features. We also create an approximation to the above algorithms that is computationally less expensive. Finally, we propose a new inter-scale method for curve segmentation and feature detection based on the motion of a curve through scale space.
Planar shape recognition by directional flow-change method1
1999, Pattern Recognition Letters
A method based on directional flow-change concept is proposed in this paper for recognizing planar shapes. This method locates critical points by evaluating directional flow-changes in terms of chain code directions along the contour. An angle string can be obtained by computing the sizes of angles at the critical points and then converting them into some special angle codes. Possible shape numbers are generated by applying a set of rules to the angle string. Consequently, the task of shape recognition is transformed to the task of comparing shape numbers. Experimental results show that our method is very effective in recognizing shapes and more efficient in computation than the curve bending function (CBF) method.
A curve bend function based method to characterize contour shapes
1997, Pattern Recognition
In this paper, a new Curve Bend Function (CBF) based method is presented to extract main features of a contour. It utilizes two quantities, one called a Curve Bend Angle (CBA), and the other known as a type coefficient of curve segment defined on a contour. The CBA measures the bending degree at each point on the contour, while the type coefficient characterizes the properties (convexity and concavity) of simple curve segments. The CBF combines the advantages of the two quantities and provides an overall description of a contour. A local maximum or minimum of the CBF corresponds to a critical point of the contour, and the peak value is the cosine of the CBA. Hence, the problem of locating critical points of a contour is reduced to the problem of finding peak points of the CBF. The sign of a peak value indicates the related CBA being an inner or outer angle. Our experimental results and comparisons between the CBF method and Rosenfeld-Johnston's method show that our method is effective and accurate in generating the description of a contour.
Effective classification of planar shapes based on curve segment properties
1997, Pattern Recognition Letters
In this paper, a pattern classifier for recognition of planar contours is developed based on the curve bend function (CBF). The CBF makes use of both the curve bend angle which measures the bend degree of a curve segment of a contour and the type coefficient which indicates whether the curve segment is convex or concave, i.e., the related corner on the contour is an inner or outer angle. The information of a contour obtained by its CBF is sufficient to represent its main features. The classifier is designed by using the properties of the CBF directly, and its training process is simpler than other kinds of classifiers, such as neural networks. Our experimental results demonstrate that the classifier is robust for planar shape classification.
Digital Geometry: Geometric Methods for Digital Picture Analysis
2004, Digital Geometry: Geometric Methods for Digital Picture Analysis

View all citing articles on Scopus

View full text

Computing dominant points on simple shapes

Pat. Recog

Computer Graphics and Image Processing

Picture syntax

Piaget and Knowledge