Hough Transform: Serial and Parallel Implementations
V. Bhatnagar, Jyoti Bala, R. Kanishka · 2011
Hough Transform(HT)[1][2] is an established technique for finding boundaries/edges in different types of shapes. Heavily used (as one of the embedded techniques) in facial features matching tools, this algorithm has its origin from the days of bubble chamber experiments used then for finding straight tracks/curves. We would like to present this algorithm from the perspective of a similar problem of finding tracks in the INO-ICAL (India based Neutrino Observatory – Iron CALorimeter) detector, which would be coming up soon in the Western Ghats. Classical Hough transform was concerned with the identification of lines in the image, but later on Hough transform has been extended to identifying positions of arbitrary shapes like circles, parabola etc. Now-a-days lots of versions have made into the Hough transform algorithm. The most common are kernel-based Hough transform and the generalized Hough transform. Although we do have good tracking code like: Kalman filter and Cellular Autometa but in order to minimize the track finding and coarse reconstruction time we can switch to Hough transform technique easily.