Various scheme and technical implementations of a vector- matrix multiplier for neuronetworks design are considered. Variants of the VMM laboratory model on the basis of the traditional optics and standard optoelectronic components are worked out. Input and output vectors are realized accordingly as radiation source and photodetector matrices; weights set with the help of a projector LCD. Possibility of an integrated optical VMM implementation as a monolithic circuit is considered; peculiarity of a planar circuit working is noted. The results of circuit elements calculations and complete VMM circuit arranging with a control block diagram are presented. This neurochip is made on the GaAs substrate by means of the traditional semiconductor technology (0.35 mkm) and takes 40 X 10 mm2 area.
The main features of scheme and technical implementation of an optoelectronic neuroprocessor for PC are considered. The neurocoprocessor provides an emulation of the one layer of any neuronetwork (with the neurons quantity to 1000 by their absolute coherence) on the basis of a vector-matrix multiplication. Worked out system is capable of discerning of very noisy images in the time, not more than three output photodetector rise times.
Single-channel optoelectronic neurocoprocessor on the base of vector-matrix multiplier with column submatrices is designed. Two models with dynamic and constant threshold are proposed for implementation. Results of pattern recognition simulation showed fast and effective convergence to the memory patterns.
Designed circuit of the large-scale high-efficiency optoelectronic vector-matrix multiplier and its components are considered in terms of geometrical optics. The results of aberration calculation are given to show the absence of cross distortions in both cascades of the system composed of the commercially available optoelectronic elements.
The comparative analysis of hologram and optoelectronic implementations of neural networks (NN) shows that at least for a small-format (N < 128) and middle-format (N equals 256...1024) optical NN the optoelectronic implementation is optimum. It is based on the original optical vector-matrix multiplier. This one contains quasi-rectangular LED-array for the N-length input vector load, projective monochrome PCs LCD for weight-matrix input, quasi-rectangular array of Si-photodiodes (or CCD), controlling PC (notebook type), and additional electronics for a linkage PC and optoelectronic neural processor. The peculiarities of such an optical system assembled from the standard optoelectronic components and providing neural processing of the middle-format images (up to 103 pixels) during one stroke of a coprocessor (10...100 ns) were examined.
This work is devoted to the problem of the further improvement of operating algorithm and technical design of the superproductive (> 1013 log.op/s) hybrid optical-digital matrix coprocessor. It is applicable to solving the overdefined linear binary equation system by means of step-by-step moving through all the solutions with the recursive use of the initial data and with allowance for degree of certainty of every equation (of the initial equation system) and of all derived solutions.
Mathematical description and algebraic interpretation of recursive algorithm in optical binary image processing are carried out. The recursive optoelectronic processor configuration is described and experimental results are presented.
The paper presents principles and design considerations of optoelectronic multi-module spectrum analyzers (direction finders) for fast processing of coherent acoustic signals. The results of theoretical investigations are discussed and the module design is considered for the systems implementing both amplitude-only and amplitude/phase processing of coherent low- frequency signals by means of radial and turn shearing interferometers.
Mathematical description and algebraic interpretation of recursive algorithm in optical binary pattern processing are carried out. The recursive optoelectronic processor configuration is described.
In the framework of the kinematic theory of three dimensional holograms
an analytical expression for the impulse response of a 3-D reflective hologram
in a paraxial approximatation has been obtained. The problem of applicability
of the transfer function concept to a 3-D reflective hologram has been studied
and it has been shown that even in a paraxial domain this hologram can he regarded as a space-invariant system only at small deviation from the Bragg condition. The analytical expression defining the transfer function of a 3-D reflectiye hologram has been derived for this case. The formulae obtained allow
to rapidly evaluate the quality of the reconstructed image.
In the framework of the coupled wave theory the dependence of the light intensity distribution in the holographic image of an extended object on the Bragg angle mismatch has been found for the phase-amplitude 3-D reflective hologram the intensity distribution across the object being arbitrary. The grey level distortions induced by this hologram have been evaluated depending on the observation point position with respect to the hologram and the reconstructing source. 1 .
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.