ABSTRACT
The roots of optical pattern recognition can be traced back to Abbe's work in 1873 [1], which led to the discovery of spatial filtering. However, optical pattern recognition was not appreciated until the complex spatial filtering of Vander Lugt in 1964 [2]. Since then, techniques, architectures, and algorithms have been developed to construct efficient optical correlators for pattern recognition. The objective of this chapter is to discuss the optical architecture and techniques as applied to pattern recognition. Basically, there are two approaches in the optical implementation, namely, the correlation approach and the neural net approach. In the correlation approach, there are two frequently used architectures, the Vander Lugt correlator (VLC) and the joint transform correlator (JTC).
