Ellipse Detection using the Hough Transform

Yuen, H. K.; Illingworth, J.; Kittler, Josef

doi:10.5244/c.2.41

Cited by 25 publications

(12 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The chord-tangent method aims to make good the deficiencies of the diameter-bisection method -specifically, its distinct lack of robustness [13]. A list of edge points is again made, and pairs of these are selected in turn and used to create votes in parameter space.…”

Section: The Chord-tangent Methodsmentioning

confidence: 99%

Object Location Using the HOUGH Transform

Davies¹

2012

Machine Vision Handbook

View full text Add to dashboard Cite

Section: The Chord-tangent Methodsmentioning

confidence: 99%

Object Location Using the HOUGH Transform

Davies¹

2012

Machine Vision Handbook

View full text Add to dashboard Cite

“…For l in this range, one may calculate the coordinates of the center of the ellipse, (x 0 (l), y 0 (l)), via (8). Note that x 0 (l) and y 0 (l) parameterize the line previously found in [18]. Observe that …”

Section: The Range Of L For Which the Conic Is An Ellipsementioning

confidence: 99%

“…If one of these conditions is not met, we use the normal vector -N k instead of N k in defining our curve. We then use the parameterizations in (11) to define the curve (x 0 (l), y 0 (l)), which we have shown parameterizes the line previously found in [18]. When considering candidate ellipses (parameterized by l) for voting during the Hough Transform step, we eliminate those with ratio of major axis length to minor axis length exceeding 4.0.…”

Section: Center-finding Experimentsmentioning

confidence: 99%

“…Specifically, the problem of ellipse detection is often computationally demanding, since an ellipse is characterized by five parameters: center position coordinates; lengths of the major and minor axes; and the orientation of the major axis. Many wellstudied approaches have been used to address this problem: 1) probabilistic methods (see, for example, [8]), which attempt to obtain all five parameters while selectively sampling the image's edge data; 2) simulated annealing approaches (see, for example, [3]); and 3) multistage methods (see [13], [17], and [18]), which detect the parameters two or three at a time during each stage.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A method to detect and characterize ellipses using the Hough transform

Bennett

Burridge

Saito

1999

IEEE Trans. Pattern Anal. Machine Intell.

View full text Add to dashboard Cite

Abstract-In this paper we describe a new technique for detecting and characterizing ellipsoidal shapes automatically from any type of image. This technique is a single pass algorithm which can extract any group of ellipse parameters or characteristics which can be computed from those parameters without having to detect all five parameters for each ellipsoidal shape. Moreover, the method can explicitly incorporate any a priori knowledge the user may have concerning ellipse parameters. The method is based on techniques from Projective Geometry and on the Hough Transform. This technique can significantly reduce interpretation and computation time by automatically extracting only those features or geometric parameters of interest from images and making exact use of a priori information.

show abstract

“…Where more than two parameters are under detection the method may involve unpractically large computation and storage requirements. To deal with this particular difficulty, a third class of methods decomposes the computation into separate stages [10], [11], [12], [13], each stage passing results to the next. The present method, the DGHT [14], [15], [16] uses information available in the relative distribution of feature points to optimise the computation of the transform.…”

Section: Parametric Transformation Is a Powerful Tool In Shape Analysmentioning

confidence: 99%

Active intelligent vision using the dynamic generalized Hough Transform

Leavers¹

1990

Procedings of the British Machine Vision Conference 1990

View full text Add to dashboard Cite

The Hough transform[l] works by grouping low level feature points (edge image data) into object specific intermediate features (e.g. line segments). This is accomplished using the low level feature points to generate information concerning all possible groupings of points within the image. The corresponding transform plane is the accumulation of that evidence. The technique is computationally intensive because evidence is generated of all possible groupings of points in the image.Previous suggested approaches may be divided into three categories. The first seeks to reduce the computational load by using information from the image to reduce the generation of evidence in the transform plane [2] [16] uses information available in the relative distribution of feature points to optimise the computation of the transform. The n parameters associated with the shape under detection are calculated using a single fixed image point and all possible combinations of that image point with sets of (n -1) other image points. Each parameter is calculated independently and requires only one dimensional accumulation of evidence. Hence if T is the resolution in transform space and n is the number of parameters under detection then use of the DGHT reduces memory requirements from T n to nT and introduces the opportunity for parallel calculation and accumulation of parameters.Peaks in the histograms will indicate the parameters associated with the most probable instance of the shape in image space of which the fixed point is a member. This information may be used to detect that shape and to remove the points associated with it from the image. The process is then repeated using the shortened list. Peak detection in the transform space normally presents difficulties. The combinatorial nature of the generation of peaks using the DGHT means that peaks are significantly higher than the background and detection is one dimensional.Global processing of the image has remained a standard feature of parametric transformation. The DGHT is sensitive to feature point density in the image and it is therefore expedient to segment the image. Uniform segmentation has been used previously [15] for a simple image of a single object. Where the image contains multiple objects or occlusion is present, segmentation cannot be fixed but needs to be adapted to the image and the type of shape under detection. The DGHT provides a feedback mechanism linking image and transform space. Decisions can be made concerning the probable instance of shape under detection, 49BMVC 1990

show abstract

Ellipse Detection using the Hough Transform

Cited by 25 publications

References 7 publications

Object Location Using the HOUGH Transform

Object Location Using the HOUGH Transform

A method to detect and characterize ellipses using the Hough transform

Active intelligent vision using the dynamic generalized Hough Transform

Contact Info

Product

Resources

About