Multi hyperbole detection on images using modified artificial bee colony (ABC) for multimodal function optimization

In this paper, a novel algorithm based on Hough transform is presented for automatic detection hyperbolas in images using a modified artificial bee colony (ABC) algorithm. Hough technique is the most common solution for detecting hyperbolas in images. This method was first introduced by Richard O. Duda for detecting lines in images [1]. The disadvantage of Hough algorithm lies in the fact that it requires large memory size and long computation time. Therefore, optimization method has been used to deal with this problem in this paper. Since optimization is used to find the best solution, the output of the algorithm will only detect one hyperbola in the image if classical optimization methods are used. In this paper, a modified ABC algorithm is presented in order to detect multiple hyperbolas in one time implementation of the algorithm. The classical algorithm is modified to a multimodal optimization algorithm. Therefore, the objective function is based on Hough method. Experiments conducted on the images made by computer (unrealistic data) showed that algorithms could detect multiple hyperbolas in one time implementation of the algorithm. Moreover, the results obtained from the conducted experiments on Noisy images that the algorithm can efficiently meet the criteria and resolve the problem.