EFFICIENT ALGORITHM FOR LEARNING OPTIMIZATION OF MORPHOLOGICAL FILTERS

An efficient algorithm based on the simulated annealing for the learni ng optimization of morphological filters is proposed. The learning sta ge is divided into two consecutive parts; the initial-learning stage f inds and fixes the most important parts of the structuring elements, a nd the precise-learning stage determines details of the rest. This met hod significantly reduces the number of trials for the modification of structuring elements. The proposed algorithm is applied to the learni ng optimization of the bipolar morphological operation, whose optimiza tion problem has not yet been investigated. It is shown experimentally that the algorithm optimizes the operator as efficiently as the conve ntional one and reduces the amount of calculation.