Hough transform (HT) has been the most common method for circle detectionexhibiting robustness but adversely demanding a considerable computational loadand large storage. Alternative approaches include heuristic methods that employiterative optimization procedures for detecting multiple circles under theinconvenience that only one circle can be marked at each optimization cycledemanding a longer execution time. On the other hand, Learning Automata (LA) isa heuristic method to solve complex multi-modal optimization problems. AlthoughLA converges to just one global minimum, the final probability distributionholds valuable information regarding other local minima which have emergedduring the optimization process. The detection process is considered as amulti-modal optimization problem, allowing the detection of multiple circularshapes through only one optimization procedure. The algorithm uses acombination of three edge points as parameters to determine circles candidates.A reinforcement signal determines if such circle candidates are actuallypresent at the image. Guided by the values of such reinforcement signal, theset of encoded candidate circles are evolved using the LA so that they can fitinto actual circular shapes over the edge-only map of the image. The overallapproach is a fast multiple-circle detector despite facing complicatedconditions.
展开▼
