Computer Vision: Detecting Parabolas Using the Hough Transform

Question

Documents have been written describing how the Hough transform can be generalized to detect shapes such as circles and parabolas. I am new to computer vision, and I think these papers are pretty tough. There is also code that makes this detection, but it is more than what I want. I was wondering if anyone could describe briefly in bullet points or pseudo-code just how Hough Transforms are used to detect parabolas in images. That would be awesome. Or, if someone knows some basic explanations on the Internet that I have not seen, that would be nice too :).

Many thanks:).

Answer 1

Interest Ask. It looks like a great resource . I have included a resume (freely quoted). Also see the Source from Mathworks at the bottom of this answer. Matlab has houghlines and houghpeaks features to houghlines you. Hope this helps.

• An edge detection algorithm, such as a Canny edge detector on an object
• Input Boundaries / Boundary Points in a Hough Transformation (Line Detection)
  • Generate a curve in polar space (radius, angle) for each point in Cartesian space (also called a battery matrix)
  • Retrieve local maxima from the drive array, for example, using a relative threshold
  • In other words, we take only those local maxima in the accumulator array whose values ​​are equal to or greater than a certain fixed percentage global maximum value.
• De-Houghing in Cartesian space gives a set of line descriptions of an image object

cs.jhu.edu: http://www.cs.jhu.edu/~misha/Fall04/GHT1.pdf

Code from Mathworks: http://www.mathworks.com/help/toolbox/images/ref/hough.html