Pattern recognition system based on photoanisotropic copies.

Abstract

In this paper, we present a method of pattern recognition based on obtaining the photoanisotropic copies of object images. Such copies are induced on the dynamic polarization-sensitive material when the amplitude image of the object formed by a spatial light modulator is transilluminated by linearly polarized light with a wavelength actinic for the used material. In this case, the distribution of the intensities over the amplitude image is converted into the distribution of photoanisotropy over the sample of a polarization-sensitive material. In the recognition process, this copy is illuminated by parallel circularly polarized nonactinic light. The distribution of elliptical polarization occurs directly behind the photoanisotropic copy, and it is reduced to a summary ellipse in the Fraunhofer diffraction region. It has been shown that the parameters of this ellipse are related to the characteristics of the object and uniquely identify the initial object. The polarization-holographic diffraction element that we developed enables determining the polarization state of a summary ellipse to obtain all the Stokes parameters in real time. The obtained results were compared with the Stokes parameters of the recognizable object in the database. The method invariance with regard to the position, scale, and rotation of the pattern were investigated. In addition, the resolution and sensitivity of this method were determined. The polarization-sensitive materials used to obtain copies are reversible, with practically an unlimited number of recording-deleting cycles. To obtain a photoanisotropic copy of another object on the same material, the previous copy should be deleted with a pulse of circularly polarized actinic light, and then the next copy can be recorded. A laboratory model of the recognition device and the appropriate software were created, and a theoretical model was developed. Databases were obtained for the images of various objects.