PCNN Edge Detection of Sintering Pellets Image Based on Hybrid Harmony Search Algorithm

Harmony search (HS) is a newly developed derivativefree and meta-heuristic algorithm mimicking the improvisation process of musicians, which has been very successful in a wide variety of optimization problems. A hybrid optimization method is proposed for self-tuning pulse coupled neural network (PCNN) parameters, a biologically inspired spiking neural network, based on harmony search algorithm and simulated annealing (SA) idea and was used to detect sintering pellets image edges automatically and successfully. The effective of the proposed method is verified by simulation results, that is to say, the quality of the sintering pellets grayscale image edge detection is much better and parameters are set automatically.