Sceptron

Sceptron

 

an optoelectronic device that analyzes complex electrical signals in the frequency range from 100 hertz to 100 kilohertz on the basis of the signals’ spectral and time characteristics. The term “sceptron” is derived from “spectral comparative pattern recognizer.”

Figure 1. Block diagram of a sceptron: (1) light source, (2) resonant array of quartz or glass optical fibers that are of different lengths (< several centimeters) and different diameters (< 100 micrometers) and are attached at one end to a common base, (3) reference mask consisting of a glass plate with a transparent or dark pattern, (4) matrix of photocells, (5) analyzing and classifying device, (6) exciter of array of optical fibers (an electrodynamic or piezoelectric transducer that converts electrical signals into mechanical vibrations)

The operation of a sceptron is based on the mechanical resonance properties of optical fibers (seeFIBER OPTICS and LIGHT GUIDE). Amplified electrical signals from a microphone or optoelectronic sensing device arrive at the sceptron input (Figure 1). The signals are fed to an electromechanical exciter and cause it and certain groups of fibers to vibrate at resonant frequencies. Light rays passing through the vibrating and stationary fibers reach a reference mask and pass through it to a matrix of photocells. On the basis of the current distribution in the photocell circuits, an analyzing device determines whether the signal being analyzed is identical to the recorded reference pattern. If the signal is recognizable, it is classified by the analyzing device.

Sceptrons appeared in the early 1960’s and are used in, for example, cryptography, medicine, and communication systems. In particular, sceptrons are used in the recognition of graphic characters and speech signals, in the analysis of information from hydroacoustic installations, in the diagnosis of heart and lung diseases on the basis of characteristic acoustic noise, and in the study of the “language” of dolphins. (See alsoPATTERN RECOGNITION.)

REFERENCES

Barchenkov, S. A. Chudesnye volokna. Moscow, 1969.
Miasnikov, L. L., and E. N. Miasnikova. Avtomaticheskoe raspoznavanie zvukovykh obrazov. Leningrad, 1970.
Galushkin, A. I. Raspoznavanie signalov na skeptronakh. Moscow, 1974.

S. A. BARCHENKOV