Approach to accurate circle detection: Circular Hough Transform and Local Maxima concept

Detecting circular objects over digital images have received considerable attention from industries for applications such as detection of pellets in pelletization plant, target detection, inspection of manufactured products etc. Several algorithms were proposed in past few years to detect circular features. One powerful approach for circle detection is the Circular Hough Transform and its variants. This article presents an algorithm which is based on CHT and Local Maxima concept. Finding one or several maxima considering different accumulators simultaneously and mapping the found parameters corresponding to the maxima back to the original image is key concept of proposed algorithm. Experiments were performed on real industrial images to validate the efficiency of proposed algorithm regarding good accuracy of detection.

[1]  Kuo-Liang Chung,et al.  An Efficient Randomized Algorithm for Detecting Circles , 2001, Comput. Vis. Image Underst..

[2]  Xiao Zhang,et al.  An Improved Circle Detection Method Based on Right Triangles Inscribed in a Circle , 2009, 2009 WRI World Congress on Computer Science and Information Engineering.

[3]  Jong-Hwan Kim,et al.  A two-step circle detection algorithm from the intersecting chords , 2001, Pattern Recognit. Lett..

[4]  Simon Just Kjeldgaard Pedersen.  Circular Hough Transform , 2009, Encyclopedia of Biometrics.

[5]  Julien Marot,et al.  Subspace-Based and DIRECT Algorithms for Distorted Circular Contour Estimation , 2007, IEEE Transactions on Image Processing.

[6]  A. Goneid,et al.  A method for the Hough transform detection of circles and ellipses using a 1-dimensional array , 1997, 1997 IEEE International Conference on Systems, Man, and Cybernetics. Computational Cybernetics and Simulation.

[7]  I. Vaughan L. Clarkson,et al.  Maximum-likelihood estimation of circle parameters via convolution , 2006, IEEE Transactions on Image Processing.

[8]  Shiu Yin Yuen,et al.  Efficient technique for circle detection using hypothesis filtering and Hough transform , 1996 .