Open Access Presidential Scholars Thesis
Sarah Rebecca Thomas
Computer vision and human vision are quite different. While a person can look at an object and easily recognize it, a computer is does not understand the the objects it is seeing. Finding axes and circles in an image is an intermediate step in designing a machine with the ability to recognize shapes and objects. Through the utilization of contour and smoothed local symmetry information, the main axes of an image are determined and may be used to facilitate shape and object analysis. A Hough transform algorithm detects the most popular lines from a set of candidate axis points. This conversion process also detects the optimal circles in an image. Through the use of Hough transform algorithms, these axes and circles are designated as being the best suited for the computer to recognize what it is seeing. These extensions bridge part of a gap between a collection of outlines and symmetric points to shape and eventual object recognition.
Date of Award
Department of Computer Science
Presidential Scholar Designation
A paper submitted in partial fulfillment of the requirements for the designation Presidential Scholar
1 PDF file (36 pages)
©1995 - Karissa E. Hobert
Hobert, Karissa E., "Efficient axis and circle detection through Hough transformation" (1995). Presidential Scholars Theses (1990 – 2006). 84.