APPLICATIONS

Pattern Recognition

By using a phase-dependant mask, our optical computing technique can readily be applied to pattern recognition. If the set of possible characters is placed in an array as the pattern for the mask, then the correlation of the mask and the incident Fourier transform will yield a bright spot at the same position on the screen as the corresponding character on the mask. One aspect of pattern recognition, namely spectral identification or recognition, has only recently been an application of optical computing methods [7]. Even by doing spectral recognition through a conventional computer, the experiment can take hours due to the thousands of computations necessary to compare the sample to the database of known spectra. In addition, this database is usually comprised solely of pure samples, so that when mixtures of many elements are observed, such analysis can prove tedious and time consuming for the software.

By utilizing spectroscopy and optical correlation for spectral recognition, we can identify materials through an efficient, cheap, and simple setup.

Optical Logic

The instantaneous nature of optical computing allows for numerous logic operations to be performed in parallel in only a few nanoseconds [5]. This is done through a coding known as shadow casting, in which light (green) and dark (red) shading correspond to ‘1’ and ‘0’ respectively. The use of color allows us to take advantage of pigment absorption to still get light through a ‘dark’ on ‘dark’ overlay (red light will still pass through a red filter). Thus, two shadow casting slides in sequence act as an AND logic operator.