Computer and Modernization ›› 2013, Vol. 1 ›› Issue (2): 22-26.doi: 10.3969/j.issn.1006-2475.2013.02.006

• 图像处理 • Previous Articles     Next Articles

Circle Detection Algorithm Based on Gradient Hough Transform and SVM

CHENG Peng1, ZHU Mei-lin1, GENG Hua2   

  1. (1. School of Management and Engineering, Nanjing University, Nanjing 210093, China;2. College of Information Engineering, Yangzhou University, Yangzhou 225127, Ghina
  • Received:2012-10-19 Revised:1900-01-01 Online:2013-02-27 Published:2013-02-27

Abstract: Hough transform is a well established method for circle detection. However, circle Hough transform is quite time-consuming. In this paper, some improvements based on standard Hough transform are employed to improve the algorithm. The 3D Hough transform is transformed into the combination of two 2D Hough transform. Gradient information is treated as the value of the voting rights in the accumulated array. SVM is used in the recognition of false circle. The experimental results indicate that the proposed approach can improve the detection accuracy and computational efficiency.

Key words: Hough transform, gradient, SVM, circle detection