Optical Pattern Recognition System Research Articles

Optical pattern recognition using multiple phase-shifted-reference fringe-adjusted joint transform correlation

We propose a novel optical pattern recognition system using multiple phase-shifted-reference fringe-adjusted joint transform correlation (MRFJTC) techniques. The MRFJTC algorithm can efficiently detect an object of interest in the input scene by producing a highly distinctive correlation peak while rejecting any and all nontarget objects in a complex background. The simple architecture of the proposed system can simultaneously recognize multiple targets of the reference class when present. The recognition performance is fast, automatic, and invariant to noise and distortions.

Optical pattern recognition with adjustable sensitivity to shape and texture

In this paper, an optical pattern recognition system with adjustable sensitivity to shape distortions and texture changes of the objects is presented. Application to a recognition task where the information of texture is the most decisive feature for a given object to be detected is provided. We apply the dual nonlinear correlation (DNC) model along with a support function acting in the frequency domain. This support function performs as an additional nonlinearity that enhances the information of some selected frequency bands related to the textural content of the target. A mathematical analysis allows the authors to show the usefulness of the proposed support function in the frame of the DNC model. The recognition system is applied to accomplish different recognition tasks involving model and real textured objects. The proposed optoelectronic correlator has been used to obtain successful experimental optical results, which are in accordance with the simulated results also provided.

Optical Pattern Recognition System Based on Parallel Spatial Filtering Using Synthetic Discriminant Function

In this paper, we propose an optical pattern recognition system based on spatial filtering using a synthetic discriminant function (SDF). In this system, the input image is multiplexed by a microlens array under incoherent illumination, and then the optical inner product of input images and SDF filters is optically conducted. The image is recognized by thresholding the optical inner product values. The present SDF filter was constructed with pseudogradation transparent filters fabricated by electron-beam lithography. We performed an experiment that involved the recognition of ten Arabic numerals and evaluated the discriminating characteristics of parallel pattern recognition. We confirm that the proposed system can function as a parallel recognition system without the need to scan images.

Optoelectronic implementation of neural networks

This article discusses the optical architectures and techniques in implementing neural networks as they apply to optical pattern recognition and optical information processing systems.

Compact optical correlator: preprocessing and filter encoding strategies applied to images with varying illumination

We present a portable, versatile optical pattern recognition system that can process 256x256 pixel images at high speed. The system is fully controlled by Windows-based software and can easily be switched between VanderLugt and joint transform operation. It combines the advantages of optics and electronics to form an advanced hybrid system where pre- and postprocessing algorithms can be varied depending on the application. Simulation and experimental results show that edge enhancement combined with binarization gives good results when applied to images where the illumination is varying.

Second-order interpattern neural networks for optical pattern recognition

By combining the autoassociative model with the heteroassociative model, we propose a second-order semi-complex IPA neural network. In this neural network, the real part or the imaginary part of a complete or partial complex vector not only addresses the corresponding part of the complex vector prestored in the neural network by autoassociation, but also addresses the corresponding imaginary or real part by heteroassociation. Owing to the introduction of the heteroassociation, the correct addressing probability is increased. Based on this neural network, a scheme for optical pattern recognition (OPR) is proposed. This scheme is composed of two cascaded optical systems; each performs a relative simple function. Compared with the matched spatial filter OPR system, the addressing results are direct and our scheme is feasible and promising.

A polymeric optical pattern-recognition system for security verification

POLYMERS that exhibit the photorefractive effect—a light-induced modulation of refractive index—are emerging as attractive materials for optical devices and processing systems1,2. Here we demonstrate one such application using our recently developed high-efficiency photorefractive polymer2. The polymer provides a nonlinear medium in which real-time all-optical image correlation, and hence pattern recognition, can be accomplished. This forms the basis of an optical security system, whereby documents are encoded with practically invisible phase masks (such masks are difficult to forge), which may then be rapidly screened to verify the authenticity of the documents. The wavelengths at which our optical system operates are compatible with commercial low-power semiconductor laser diodes, and the system can be integrated into a compact device at low cost.

Nonlinear techniques in optical pattern recognition: introduction by the feature editors

This feature of Applied Optics is dedicated to research in optical pattern-recognition techniques that utilize nonlinear devices and algorithms. Nonlinearities are employed in a variety of ways in optical pattern recognition and play an important role in both hardware implementation and software development for optical pattern-recognition systems.

Real-time pattern recognition system based on a bipolar winner-take-all model with a threshold

A threshold is added to recognize an input pattern that has too few similarities to be one of the memory patterns. A real-time pattern recognition system based on the improved model is given. The unipolar optical pattern recognition system has all the characteristics of a fully bipolar winner-take-all model and has illumination invariance. A liquid-crystal spatial light modulator is used as a real-time input device, and a mask and a lens array perform the threshold and weighted sums of the input pattern. Some experimental results are also shown.

Correlation peak degradation for out-of-plane rotation in an optical pattern recognition system

Optical correlation is sensitive to out-of-plane rotations of the object in the input image. Such an out-of-plane rotation causes the correlation peak to degrade. Here, the degradation of this correlation peak is investigated and a model is proposed to explain its behavior for arbitrary convex objects. First an analytical expression is derived for the peak degradation caused by out-of-plane rotations of flat objects, which is equivalent to a scaling in one dimension. This analytical result compares favorably with experimental results obtained from computer simulation. The 1-D theoretical model is then generalized to explain the behavior of the correlation peak for out-of-plane rotations of arbitrary convex objects. Experimental results for the latter case are also provided.

An OPTICAL PATTERN recognition system for validation & security verification

an OPTICAL PATTERN recognition system for validation & security verification

Phase-encoding input images for optical pattern recognition

The images introduced to an optical pattern recognition system are normally an amplitude representation of the original intensity images. We define a method of encoding intensity images as phase-only images, and we demonstrate that the performance of the correlation is generally substantially better than that of the correlation performed with amplitude inputs. Energy in the correlation peak is about a factor of 20 greater for the phase-encoded input image method. These performance and energy advantages are particularly useful when designing filters that incorporate various degrees and types of invariance. We give examples and numerical results for the correlation process with ternary and 15-phase-state plus zero-amplitude filters with varying spans of invariance. We suggest practical implementations for this technique.

Optical pattern recognition system based on a winner-take-all model of a neural network

A three-layer neural network implemented by optical and electronic techniques for pattern recognition is described. The principle, the architecture, and the experimental results of the hardware system for pattern recognition are presented. A multiplex hologram is used as the interconnections between neurons of the input layer and the middle layer. The winner-take-all (WTA) operation of the intermediate layer is performed by an electronic circuit. The interconnection between neurons of the middle layer and the output layer is simply implemented with an optical transparency. By changing the patterns on the transparency, this system can be used for both auto- or heteroassociative memories. In addition to noise and partial hindrance tolerances, it also has illuminance invariance because of the WTA layer of the system.

Optical pattern recognition of letters by a joint-transform correlator using a ferroelectric liquid-crystal spatial light modulator

The use of a joint-transform correlator (JTC) as an optical pattern-recognition system for letters has been studied. The system incorporates a ferroelectric liquid-crystal spatial light modulator (LAPS-SLM). The LAPS-SLM has a high contrast ratio and a fast response time, permitting the JTC to recognize more reference letters simultaneously than a classical correlator can.

On the number of allowable classes in an optical multichannel pattern recognition system

In a recent publication the operation of spatial filters in the presence of noise was analyzed in terms of detection probability and discrimination. The analysis is extended here to multichannel detection systems for multiclass pattern discrimination. The results are employed to determine the number channels allowed before a given error probability may be exceeded. Taking into account the discrimination limitations of the optical system, a case study demonstrates the superiority of a specific synthetic spatial filter over the conventional matched filter.

Resolution limits of holographic mapping filters for geometric coordinate transformations

A recently proposed position, rotation, and scale-invariant optical pattern recognition system using a holographic logpolar mapping filter is analyzed with respect to resolution and accuracy. The limits are set by the resolution of the mapping filter, which must have a bandwidth proportional to the square of the input bandwidth. With 1800 x 1800 pixels in the computer-generated mapping filter, we get a maximum input field of 24 x 24 pixels and an angular resolution of 4.80 degrees . Determination of the position of the input object is possible with at best 50% relative accuracy, and, in practice, multiobject capability is highly questionable. Numerical simulations and experimental evidence support the resolution limits found.

A multichannel optical correlator

A multichannel parallel optical pattern recognition system is proposed. The synthesis of a large capacity matched filter is discussed. The technique uses a phase mask and telescopic arrangement for the mass filter construction. The processing capacity of the proposed optical correlator, which is over 400 times of a single channel system, is evaluated.

Large-memory real-time multichannel multiplexed pattern recognition

The principle and experimental design of a real-time multichannel multiplexed optical pattern recognition system via use of a 25-focus dichromated gelatin holographic lens (hololens) are described. Each of the 25 foci of the hololens may have a storage and matched filtering capability approaching that of a single-lens correlator. If the space–bandwidth product of an input image is limited, as is true in most practical cases, the 25-focus hololens system has 25 times the capability of a single lens. Experimental results have shown that the interfilter noise is not serious. The system has already demonstrated the storage and recognition of over 70 matched filters—which is a larger capacity than any optical pattern recognition system reported to date.

Optical Feature Extraction Via The Radon Transform

We show that by first measuring the line integrals of a two-dimensional picture f(x, y) via the Radon transform, certain feature-extraction operations useful in pattern recognition may be computed very easily; that is, the computational overhead for these feature-extraction operations is much less in the Radon space than in the direct space. In particular, we consider the following features: (1) moments of f(x, y) invariant to translation, rotation, geometric scaling, and linear contrast scaling; (2) two geometric features, polar projections and convex hull, that have similar invariance properties; (3) the Fourier power spectrum of f(x, y) integrated along radial piewedge and annular bins in the Fourier space; and (4) the Hough transform for detection of straight edges. Much of the motivation for this work lies in its implementation as an optical pattern recognition system. We show an optical-digital hardware implementation for the rapid computation of both the Radon transform and the feature extractions. By rapid, we mean that the transform and feature-extrac-tion operations on a 512 X 512 image may be accomplished at video rates (1 /30 s per image). We point out advantages of this system over more traditional optical pattern recognition schemes that rely on the use of Fourier optics. Some experimental results are shown.

Hybrid incoherent optical pattern recognition system

A pattern recognition system that uses incoherent spatial filtering to recognize images directly from a narrowband phosphor television monitor is described. Images of real objects are captured with a television camera. These images are then edge-enhanced electronically and displayed on the TV monitor. The monitor output is used directly as the input to a holographic correlator. An optical multichannel analyzer at the correlation plane is used to analyze the shape of the correlation function and to determine the position of its peak. Experimental results agree well with theory. Concepts for handling rotation, aspect angle, and scale variations of the input are discussed.