Traditional Lines and Circles detector
Jimut Bahan Pal · 2020
The traditional method for extracting features from image is accomplished by Hough Transform. When an image is represented mathematically, it becomes easy to perform computation according to a certain algorithm and detect the required features from the image. The presence of such features is determined by a voting procedure in a parameter space, from which the selected features are obtained as local maxima by creating a certain threshold. This procedure for simple detection of lines was patented by Paul V.C. Hough in 1962. There has been various advancement in this field for detecting lines and circles, all of them are based on the parent Hough Transform way to vote for the corresponding features. The various ways of detecting features and the mathematics behind them is discussed in this report.