Non-parametric dominant point detection

doi:10.1016/0031-3203(91)90004-O

Pattern Recognition

Volume 24, Issue 9, 1991, Pages 849-862

https://doi.org/10.1016/0031-3203(91)90004-O Get rights and content

Abstract

A new method for detecting dominant points is presented. It does not require any input parameter, and the dominant points obtained by this method remain relatively the same even when the object curve is scaled or rotated. In this method, for each boundary point, a support region is assigned to the point based on its local properties. Each point is then smoothed by a Gaussian filter with a width proportional to its determined support region. A significance measure for each point is then computed. Dominant points are finally obtained through nonmaximum suppression.

Unlike other dominant point detection algorithms which are sensitive to scaling and rotation of the object curve, the new method will overcome this difficulty. Furthermore, it is robust in the presence of noise. The proposed new method is compared to the Teh-Chin algorithm (see C. Teh and R. T. Chin, IEEE Trans. Pattern Anal. Mach. Intell.11, 859–872 (1989)) in terms of the computational complexity, the approximation errors and the number of detected dominant points of an object curve.

References (17)

N. Ansari et al.
On detecting dominant points
Pattern Recognition
(1991)
U. Ramer
An iterative procedure for the polygonal approximation of plane closed curves
Comput. Graphics Image Process.
(1972)
P.V. Sankar et al.
A parallel procedure for the detection of dominant points on a digital curve
Comput. Graphics Image Process.
(1978)
F. Attneave
Some informational aspects of visual perception
Psychol. Rev.
(1954)
L.S. Davis
Understanding shape: angles and sides
IEEE Trans. Comput.
(1977)
L.S. Davis et al.
Detection of step edges in noisy one-dimensional data
IEEE Trans. Comput.
(1972)
J.G. Dunham
Optimum uniform piecewise linear approximation of planar
IEEE Trans. Pattern Anal. Mach. Intell.
(1986)
H. Freeman et al.
A corner-finding algorithm for chain coded curves
IEEE Trans. Comput.
(1977)

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

Cited by (81)

Dominant point detection based on suboptimal feature selection methods
2020, Expert Systems with Applications
This paper presents a viable alternative solution for dominant point detection predicated on the comparison of suboptimal feature selection methods. Suboptimal feature selection methods are utilized as standard criteria to identify dominant points. Considering that all of the combinations of points comprise many sets, an algorithm that eliminates some of them is affirmed and illustrated. The sequential backward selection, sequential forward selection, generalized sequential forward selection, generalized sequential backward selection and plus l-take away r selection methods are performed on the remaining points to extract the dominant points. The simulation results exhibit that this method is significantly more effective and efficient in comparison to other proposed methods.
A tangential approximation algorithm for measured data reduction of blade section curves
2018, Measurement: Journal of the International Measurement Confederation
Citation Excerpt :
They can be classified into three categories: (1) polygonal approximation method [3–6]. ( 2) dominant detection method [7–23]. ( 3) spline curve approximation method [24–35].
In order to reduce the measurement time of blade in the machining process, a data reduction algorithm is proposed in this paper. The proposed algorithm first selects a set of starting points on section curves. And then change-points are found according to its tangent properties for different segments divided by the starting points. In each segment redivided by the change-points, a redefined criterion uses a variable distance between the approximating line and section curve as objective function to ensure the proposed algorithm reserves more points where the curvature of the curve is large. This makes the achieved polygon accurately retains both the shape and details of section curves. When compared with other polygonal approximation algorithms, the proposed algorithm gives consideration to both computation time and compression rate, obtains a better approximation error distribution. Experimental results show that the proposed algorithm is stable and useful in actual machining process.
Effective and efficient contour-based corner detectors
2015, Pattern Recognition
Corner detection is an essential operation in many computer vision applications. Among the contour-based corner detectors in the literature, the Chord-to-Point Distance Accumulation (CPDA) detector is reported to have one of the highest repeatability in detecting robust corners and the lowest localization error. However, based on our analysis, we found that the CPDA detector often fails to accurately detect the true corners when a curve has multiple corners but the sharpness of one or a few of them is much more prominent than the rest. This detector also might not perform well when the corners are closely located. Furthermore, the CPDA detector is also computationally very expensive. To overcome these weaknesses, we propose two effective and efficient corner detectors using simple triangular theory and distance calculation. Our experimental results show that our proposed detectors outperform CPDA and nine other existing corner detectors in terms of repeatability. Our proposed detectors also have a relatively low or comparable localization error and are computationally more efficient.
A robust method for scale independent detection of curvature-based criticalities and intersections in line drawings
2015, Pattern Recognition
A novel two-phase iterative method is proposed to identify curvature based criticalities and intersections in line drawings. The method is based on evaluating a field obtained from the image via diffusion using adaptively changing ellipse-shaped analysis windows. The deviation of the average field strength values within the analysis windows from that of an external reference field is used as a metric to evaluate the criticality of a region. The external reference field is computed from an image of a straight line. The experimental results depict that the method is effective in detecting curvature based criticalities and intersections, even for noisy and disconnected drawings as well as drawings drawn with a variety of brush characteristics, such as glass, ocean, ripple, scatter, stamp, strokes effects. Our method can be employed as a part of a design grammar interpreter.
A novel framework for making dominant point detection methods non-parametric
2012, Image and Vision Computing
Most dominant point detection methods require heuristically chosen control parameters. One of the commonly used control parameter is maximum deviation. This paper uses a theoretical bound of the maximum deviation of pixels obtained by digitization of a line segment for constructing a general framework to make most dominant point detection methods non-parametric. The derived analytical bound of the maximum deviation can be used as a natural bench mark for the line fitting algorithms and thus dominant point detection methods can be made parameter-independent and non-heuristic. Most methods can easily incorporate the bound. This is demonstrated using three categorically different dominant point detection methods. Such non-parametric approach retains the characteristics of the digital curve while providing good fitting performance and compression ratio for all the three methods using a variety of digital, non-digital, and noisy curves.
A method for recognizing overlapping elliptical bubbles in bubble image
2012, Pattern Recognition Letters
Direct imaging technology is an effective and convenient method for the estimation of bubble size distribution (BSD). However, overlapping bubble has an influence on BSD when gas holdup is more than 1%. In this paper, we present a new method of overlapping elliptical bubble recognition to determine bubble size. The method mainly includes two steps: contour segmentation and segment grouping. Contour segmentation is on the assumption that the concave points in the dominant point sequence are always the connecting points, and segment grouping is mainly based on the average distance deviation criterion. Both simulated images and real bubble images are used to evaluate this new method. The results show that it is effective in the recognition of overlapping elliptical bubbles and have a potential in other elliptical object recognition. In the last, two methods are used for BSD estimation. It is found that the bubble size (such d10 or d32) estimated by the ignore method is slightly smaller than that estimated by the recognition method.

View all citing articles on Scopus

View full text

Non-parametric dominant point detection

Abstract

Pattern Recognition

Comput. Graphics Image Process.

Comput. Graphics Image Process.

Some informational aspects of visual perception

Psychol. Rev.

Understanding shape: angles and sides

IEEE Trans. Comput.

Detection of step edges in noisy one-dimensional data

IEEE Trans. Comput.

Optimum uniform piecewise linear approximation of planar

IEEE Trans. Pattern Anal. Mach. Intell.

A corner-finding algorithm for chain coded curves

IEEE Trans. Comput.