An efficient algorithm based on the simulated annealing for the learning optimization of morphological filter is proposed. The learning stage is divided into the following two consecutive parts; the initial-learning stage finds and fixes the most important parts of structuring elements, and the precise-learning stage determines details of the rest only. This method significantly reduces the amount of trials for the modification of structuring elements. The proposed algorithm is applied to the learning optimization of the bipolar morphological operation, whose optimization problem has not been investigated yet. It is shown experimentally that the algorithm optimizes the operator as efficiently as the conventional one and it reduces the amount of calculation.
mathematical morphology, nonlinear filter, image processing, simulated annealing, neural networks