ABSTRACT

The impetus of using photonics technologies and devices for pattern recognition started with a pattern correlator developed by Vander Lugt in 1963. With the progress of photonic devices for interfacing data and images, new techniques and algorithms slowly emerged. Photonic implementation for the recognition of pattern can basically be classified into two approaches. The first one is the correlation approach based on photonic implementation of Fourier-domain matched filters, or space-domain linear and nonlinear filters. The second approach is based on the implementation of various artificial neural network (ANN) models using photonic devices and architectures.