Abstract
Inspired by the principles from neuroscience, a spiking neural network is proposed to perform characters image processing. The main units of spiking neural network are spiking neurons which utilize inter spike time intervals as sources of information. Research on spiking neural networks has gained momentum in the last decade due to their ability to mimic biological neural network signals and their efficient computational capabilities. A supervised learning algorithm for spiking neural networks which receive input spike trains (presynaptic inputs) is proposed. In this algorithm, learning is performed in two stages: mapping of the input spike train into a spatio-temporal pattern; and use of a simple learning criteria to change synaptic weights. The proposed algorithm is then used to classify various character images.