The trends in the development of various holographic memory systems and recording materials and their prospective applications are discussed.

Recent advances in the development of a new light-sensitive media for recording, storage of optical information are discussed. Particular emphasis has been placed on the self- developing photochemical materials including media with the luminescent read-out as well as the use of non-linear optical effects for writing optical information.

An optical implementation of the Hough transform (HT) based on a matrix of 64 X 64 four-level phase computer generated holograms is described. The HT holograms are designed using a novel algorithm that combines the high speed of convergence of iterative Fourier algorithm with the high precision of the direct binary search algorithm. The matrix of holograms was fabricated using standard microfabrication techniques: E-beam lithography, photolithography and chemically assisted ion beam etching. The fabricated elements were characterized experimentally and used in the HT processor, demonstrating practical application examples as real-time straight line, ellipse, circle detection, and other pattern recognition tasks.

We consider an interesting domain of diffractive optics, namely the physics and technology of optical interface with subwavelength-structured features. Such interfaces act as effective media, which may, e.g., simulate the operation of multilayer film stacks and often feature anisotropic optical properties. The principles of the analysis and design of these interfaces are covered, and several prominent fields of application are described. The challenge of fabricating subwavelength-structured interface by low-voltage electron beam nanolithography is considered.

This article addresses the development and progress in the rapidly growing area of optical pattern recognition. In optical pattern recognition there are two basic approaches; namely, matched filtering and associative memories. The first employs optical correlators and the later uses neural networks. This paper reviews various types of optical correlators and neural networks as applied to real-time optical pattern recognition for which some of the recent advances are included.

The design and realization of a compact volume holographic memory using phase-coded multiplexing is described. We present a system capable of storing up to 480 data pages with a resolution of 640 X 480 bits in a single interaction region of a photorefractive LiNbO₃ crystal. A characterization of the key components of the system is given. First experimental results of high-resolution analog image storage as well as the application of the system to real-time optical image processing are presented.

The multilayered structure of the data storage is considered for the 2D multichannel holographic memory. The influence of the number of layers on the signal to noise ratio at the photodetector is estimated for the different diffraction efficiencies and number of channels. The comparison is made with the 1D case. The use of the random distribution of the initial phases in the channels during the recording is proposed for the diminishing of the intermodulation noises during the reconstruction. It is shown that the 2D holograms are more preferable than the 1D, as they enable greater capacity increase by adding more layers for the same requirements to the readout quality.

The important role is theoretically shown of processes of transformation of noise components of field, nonlinear in registered object field, into informational ones at reconstruction of information in system, composed of thin linear off-axis hologram (with one or several registered images) and wide-aperture phase-conjugate mirror, working in linear regime. It is shown, that for associative reconstruction of information the second, being formed at the repeat scattering by hologram, order of diffraction of the system may be used, at scattering into which noises reconstruct whole object fields.

Cross-talk noise of the holographic data storage in the superposed volume holograms is discussed. For the various methods of data encoding is determined the signal - cross- talk noise ratio depending on the diffraction efficiency of holograms, their number and amount of stored data.

The opportunities of the creation of the high-speed (to 10⁹ bits/s) holographic disk memory having a diffraction-limited capacity (to 2 - 40 Gbytes per 1 disk) are shown. This memory is based on the 3D-recording of the multiplied microholograms on the photothermoplastic carrier. New abilities of recording, storage and processing systems based on such memory are considered.

Hologram as a fuzzy relation of two subsets is treated. Fuzzy number as a Fourier-transform of the distribution in the plane of hologram is determined. To generate monoid of fuzzy numbers, convolution as an inner composition law is used. Realization of extended ariphmetic operations by holography is discussed.

Dynamic phase holograms recording in poly(methylmethacrylate) glasses containing anthracene derivatives and photochemically neutral component (residual solvent) is investigated and photoinduced diffusion of neutral molecules is proposed as a basis of the hologram evolution. The reversible recording under high temperatures and a sensitivity of the diffraction efficiency kinetics to external stresses are demonstrated. Some discrepancies between the experimental data and the offered model are discussed.

Peculiarities of the recording of dynamic phase holograms in media containing bacteriorhodopsin and problems connected with these media application for optical memory are discussed in the paper. The effect of the transitional energy exchange during recording of phase holograms in bacteriorhodopsin under the condition of unequal intensity of writing beams was observed. This effect is explained by moving of dynamic holograms during their recording.

The method of extracting images recorded in the medium that bleaches under exposure to incoherent reading light has been studied theoretically. It is shown that layer-by-layer reading the information may be realized when temporal filtration of the light field with low-frequency cut-off is applied. The most fruitful application of the technique is reconstructing time sequences of images.

The prospects for creating digital optical-memory systems based on two-photon amplitude data writing with the use of femtosecond light pulses and fluorescent data reading are analyzed. Phase and amplitude approaches to the implementation of 3D optical-memory devices are compared. The main results of investigations carried out at the International Laser Center of Moscow State University in 1996 are presented.

Investigations on 3D optical data recording in polymethylmethacrylate with spiropyran molecules by means of femtosecond Ti:Sapphire laser have been performed. The possibility for modification of polymer optical properties in volumes with sizes of 2 micrometers and recording time of < 1 ms has been demonstrated. A high image contrast of the recorded data allows to use amplitude methods of reading by both transmitted and reflected light, including parallel readout at a high speed.

The possibilities of optimizing data writing and reading in devices of 3D optical memory using photochromic materials are discussed. We quantitatively analyze linear and nonlinear optical properties of induline spiropyran molecules, which allows us to estimate the efficiency of using such materials for implementing 3D optical-memory devices. It is demonstrated that, with an appropriate choice of polarization vectors of laser beams, one can considerably improve the efficiency of two-photon writing in photochromic materials. The problem of reading the data stored in a photochromic material is analyzed. The possibilities of data reading methods with the use of fluorescence and four-photon techniques are compared.

Specialized setup for implementation of principles of two- photon writing and fluorescence reading in 3D optical memory systems based on multilayer polymer structures containing fluorescing photochromes has been developed. Modeling of the recording, readout and erasing processes have been realized. This prototype is also intended to test the photosensitive materials.

Rigid nonlinear field structures in a Fabry-Perot interferometer with nonlinear Kerr medium are investigated both theoretically and numerically. We discuss in details the conditions of the existence of localized rigid structures. The theoretical description of such stationary structures in monostable case is provided. A new phenomenon- -self-building of 1- and 2D structures--is discussed.

Applications of memory function utilizing spatial light modulators are emphasized. The optical association using analog memory of MSLM, the displacement meter using binary memory of FLC-SLM and the bistable optically addressable memory system using PAL-SLM are reviewed.

An electrically addressed spatial light modulator (SLM), which is non-pixellated and has a high diffraction efficiency, has been developed by combining an optically addressed liquid crystal SLM with a CRT. It has been confirmed that the designed SLM has excellent performance for CGH (computer generated hologram) reconstruction.

An aperture information capacity is an important characteristic of acousto-optical devices which determine the rate of addressing and recording of information into optical memory systems. The investigation of information capacity distribution along the aperture allows to optimize the acousto-optical device design parameters respectively to optical memory systems.

Three types of scanners, designed for linear deflection and using acoustic wave with a linear frequency modulation are considered. High-speed scanner with accompaniment of `a traveling acoustooptical lens'. The scanner on the basis of traveling acoustooptical lens. The scanner working in the television standard. Basic performances of scanners and the results of the experimental researches of them are presented.

The paper presents logical and physical description of the family of 3D (multilayer) electrooptic matrices with the programmable structure. Any matrix simulates the functioning of not a single digital circuit, but the functioning of a whole set of virtual digital circuits at once. With these matrices one can build digital devices executing algorithms of massive data processing.

In the report it is shown a possibility of reduction drive power of acoustooptical components, created on the base of monocrystal paratellurite (TeO₂). Geometry of acoustooptic interaction are specified, taking into account peculiarities of propagation of a slow shear acoustic wave. Results of experimental researches on creation of components with low drive power are presented.

We derive an analytical expression for the variance of the timing jitter of a soliton transmission system using sliding-frequency guiding filters, taking into account the third-order filter term. We show that the variance of timing jitter is significantly increased by the sliding action. As a consequence of the third-order filter contribution, the timing jitter is lower in a system with down-sliding than it is in a system with up-sliding at the same sliding rate. Consequently, in what concerns the timing jitter, it is advantageous to use down-sliding.

The key component for the information processing systems exploiting unlimited circulation or transmission of digital optical trains through fiber waveguides is optical pulse regenerator. Two types of travelling-wave structure based lumped regenerators are recognized and in both the cases the basic properties of structures provide the stabilization effect relative to all over the parameters of outlet picosecond pulses without regard to perturbations of incoming pulses. We develop the method of comprehensive scalling that is associated with the needed properties of multilayer structures as well as with behavior of pulses. The pursued consideration shows that multilayer semiconductor structures make it possible to reshape, and in particular to recover, all-optically both the fundamental solitons and the guiding-center solitons, as the case requires in view of their application to information processing systems.

The results of spectropolarimetric researches of fiber waveguides in a IR-range of wave lengths (0.8 - 1.6 microns) are submitted. The influence of the sizes of a core and length of the fiber waveguide on a degree of polarization of radiation on its output is shown under excitation of waveguide by both thermal, and laser sources of light. The results of researches allow to optimize a choice and modes of operations of optical and optoelectronic components for construction of neural networks of information processing.

A high-speed laser scanning system based on a cascade acoustooptical deflector is suggested and developed. Real- time generation of volumetric images with an information content of 10⁵ pixels per frame has been implemented experimentally; it exceeds by an order of magnitude and more the possibilities of the known laser acoustooptical systems.

We constructed optical systems for learning neural networks with 2D structure sing Selfoc microlens arrays. On the systems, pattern recognitions of typed alphabetic characters which were directly detected with a CCD camera were realized. Liquid crystal devices, an electron-beam addressed spatial light modulator and a Pockels readout optical modulator were used for displaying weight matrices of the neural networks. The weights were renewed according to the random search algorithm or the delta rule with error signals calculated optically. The 2D structure for image processing can be implemented with large scale networks that consist of several thousands input neurons.

We report on ongoing work with a compact all-optical recurrent neural network with 16 X 16 channels and 256 X 256 reconfigurable interconnects (weights). We will present the optical setup and report on experimental work with the system and its building blocks. The microlens-based setup shows excellent imaging properties and easy alignability. After optimizing the setup, losses could be realized by more than an order of magnitude. The system performance is currently limited by inhomogeneities of the thresholding device.

A 2D optical neuron array and its bipolar optical interconnection system are presented. Each neuron in the array consists of a liquid crystal cell and two photodetectors that detect the encoded positive and negative light inputs respectively. The transmissivity of the LC cell that represents state of the neuron is determined by the subtraction of lights that are incident onto the two photodetectors. An optical fully interconnected neural system using the neuron array and a lenslet array for bipolar interconnection is proposed.

A new neural network architecture with associative properties is proposed. Elements of every stored image obtain a special mark after local connections training. `All to all' connections are replaced by connections with a special Central Generator. It allows to unite all elements with the same mark into the whole image. The new architecture leads to a radical simplification of the connection system and therefore to the possibility of a large associative memory creation. Analog and digital optoelectronic versions are discussed.

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.

Some problems of creation of medical expert systems and the ways of their overcoming using artificial neural networks are discussed. The instrumental system for projecting neural network algorithms `Neural Architector', developed by the authors, is described. It allows to perform effective modeling of artificial neural networks and to analyze their work. The example of the application of the `Neural Architector' system in composing an expert system for diagnostics of pulmonological diseases is shown.

Model of Artificial Vision is discussed. The task is to find the part of reference pattern that presents in the object image. Condition is the task has to be solved in the framework of the Classifical Approach to the image understanding, i.e. without formal description of patterns. In two-layered bidirectional optical neural network image layer as a comparison layer is realized. Function of selective attention as filtering function in the correlation layer is involved. Energy function for the NN is defined as scalar product of object image and formed by NN expectation.

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.

This paper is devoted to the study of associative memory in the networks of N coupled nonlinear oscillators interacting via complex-valued weights. Exact solutions relating to the structure of attractors have been obtained. The complete solution to the systems of two oscillators and the structural portrait of the governing dynamical system have been obtained. It is shown that homogeneous closed chains of oscillators play important role in the context of phase associative memory problems. Qualitative description of the memory in the closed chains of N oscillators is given for arbitrary N, and rigorous solutions for N <EQ 6 are illustrated. The networks considered admit electronic, nonlinear optical and optoelectronic implementations. The background of some of them is under development.

It is proposed a new neural network which can be very useful for solving of complex pattern recognition problems, in particular recognition of images in 3D optical-memory systems.

The recognition of targets independently of orientation has become fairly well developed in recent years for in-plane rotation. The out-of-plane rotation problem is much less advanced. When both out-of-plane rotations and changes of scale are present, the problem becomes very difficult. In this paper we describe our research on the combined out-of- plane rotation problem and the scale invariance problem. The rotations were limited to rotations about an axis perpendicular to the line of sight. The objects to be classified were three kinds of military vehicles. The inputs used were infrared imagery and photographs. We used a variation of a method proposed by Neiberg and Casasent, where a neural network is trained with a subset of the database and a minimum distances from lines in feature space are used for classification instead of nearest neighbors. Each line in the feature space corresponds to one class of objects, and points on one line correspond to different orientations of the same target. We found that the training samples needed to be closer for some orientations than for others, and that the most difficult orientations are where the target is head-on to the observer. By means of some additional training of the neural network, we were able to achieve 100% correct classification for 360 degree rotation and a range of scales over a factor of five.

Optoelectronic neural networks are investigated from the viewpoint of invariant pattern recognition. Coherent and incoherent methods are used for filtration, selection and analysis of input image characteristics. Requirements on the optoelectronic components of neural networks are evaluated.

A neural network (NN) for recognition of a few image classes is considered. The method of recognition in the initial and transformed space of the features, where the training set is based on dispersion analysis, is investigated. Simulation results of the recognition process by type (alpha) NN are presented for the images of objects in the initial and transformed space of the features.

Analysis of disorders in microhemocirculation appearing during development of various genesis pathological processes allows to manifest their start-up and pathogenic mechanisms. Numerical estimation of capillary and arteriolar-veinular bed parameters gives a possibility to predict development of these states. Automatically controlled estimation of even simple parameters of microhemocirculation bed, such as vessel length in the eye conjuctiva, their amount, size, winding, etc., allows to speak about a new method in the express diagnostics of computer capillarometry. The present work is devoted to the methods of automated determination of microvessel length. The problem is solved by application of one-dimension Kohonen's networks for adaptive uniform piecewise approximation of capillary images. Kohonen's units are the approximation points, and their position is set during the self-organization of the neural network.

The comparison of recently suggested algorithms of fast approximate Karhunen-Loeve (KL) transform when applied to pattern recognition problem is presented. It is known that adequate signs selection is still the problem. In the paper we consider several algorithms that can perform fast and qualitative signs selection. Among them are wavelet based algorithm of approximate KL transform, 2D algorithm of KL transform, algorithm with projection into proper orthogonalized basis, and real time algorithm of KL transform. To compare quality of the algorithms they are applied to human faces recognition problem. The efficiency of the all abovementioned algorithms is demonstrated.

A correlator-interferometer for the phase objects real time recognition has been developed. The certain conditions under which it may operate successfully, have been examined.

The procedure of selection of the objectlike areas from the images of the basis of geometrical features is considered in this paper. The local anisotropic features of the images are used as the geometrical features. The problem of selection of the objectlike areas is considered here as the problem of extraction of image areas with the properties are close to the properties of sample subimage (object), i.e. it is the subdivision of the image into `background' and `object' parts. General sense of the approach presented here is transformation of image to the such kind, when the background and objectlike areas of the image are maximally divided, and determination of rule for division of the areas. The localglobal strategy of the image analysis, so- called effect of a rebound, enables to determine parameters of the description of the image, which are optimum from this point of view. The description of the areas, we are interested in, is defined by the appropriate samples, the origin of which is determined by the given problem. Real, graphic and synthesized images can be used as samples. The given technique is based on properties of image only, and it enables to select the areas with a given configuration of the standard objects. Self-organizing noninstructable procedure of selection can be considered as a stage of object recognition.

In this paper we will show that temporal and spatial coherence can be exploited from an incoherent source for complex amplitude processing. The method is by dispersing the spectral content and spatially encoding the incoherent source such that high temporal and desired spatial coherence can be obtained. To confirm our argument, experimental demonstrations are provided.

A new all-optical nonlinear fringe-adjusted joint transform correlation technique is proposed. The proposed scheme is based on incoherent-to-coherent conversion with the erasure of a four-wave mixing arrangement in a real-time photorefractive media. This technique is very fast because it generates the correlation output in one step and all intermediate processing is performed in the optical domain. Furthermore, this technique eliminates the problems due to nonlinearities associated with the square law devices and spatial light modulators used for recording the joint power spectrum. Computer simulation results show that the proposed joint transform correlator yields superior performance when compared to the alternate joint transform correlator architectures.

The main advantages of a neural networks are the flexibility of architecture and ability to training. It allows to perform data processing even when the processing procedure can not be present by the known function. The neural networks based on optical elements allow real time operation rather than electronic neural networks. We offered system, based on neural networks, capabling to solve a broad class of heterogeneous problems and adapting to input information flow. The system are based on neural networks of two types: (1) a single-pass network of perceptron type and (2) recognize network of Hopfield network type. The neural network of the type (1) is realized by holographic scheme and experimental checked when processing the signals fiber- optic measuring network.

Basic functional characteristics of optoelectronic 3D chips, technological and architectural principles based on them for optoelectronic computing 3D platform building are presented. These platforms and optoelectronic chipset itself may be considered as the quite suitable architectural basis for neural nets.

This paper is devoted to implementation of the some original image processing algorithms using cellular neural networks on multi-valued and universal binary neurons. High-effective multi-valued filters, their implementation and application to the noise removing and frequencies correction are considered. Algorithms of the edge detection and impulse noise filtering, and their implementation are also considered. Application of the networks from multi-valued neurons to image recognition and to extrapolation of the image orthogonal spectrum to the highest frequencies domain is also presented.

A neural network for contour pattern classification using the projective data compression technique is considered. By way of test, the method was applied to the recognition of handwritten characters. Capabilities of the method are demonstrated. A prospect of implementation is discussed.

The continuous logic `equivalental' models of Hamming neural networks with adaptive-correlated weighting and multiport associative memory based on equivalence operation of neural logic are considered. The models for simple network with weighted correlation coefficient, for network with adapted weighting and double weighting and their system equivalental functions are suggested. The models require calculations based on two-step algorithms and vector-matrix procedures with the normalized equivalence operation. Modified equivalence models of neural networks and associative memory for space-invariant 2D pattern recognition are proposed. Possible variants of the models implementation are considered. Neural networks architecture for invariant 2D pattern recognition consist of equivalentors, every of which replace two correlators.

The method of coding of input binary data based on their transformation to binary-sign-bit notation data by special algorithm is proposed. The method enables to eliminate through carries and reduce difference in times of fulfillment of various arithmetic operations thereby increasing considerably the speed of calculations in optoelectronic digital computing systems.

Natural biomolecular information processing utilizes conformational (shape-based) interactions that support capabilities complementary to those of conventional computers. Optical technologies provide a means for interfacing these inherently different modes of processing to achieve hybrid systems with unique computational synergies. Part one of this two part paper addresses the conformational, kinetic, structural, and dynamic features of biomolecular systems that are pertinent to the processing of optical signals. Part II describes a simulation tool for studying the interaction of these factors and considers how this tool could in the future to be used to design biomolecular devices capable of processing optical signal patterns.

A highly flexible simulation tool has been developed to study biomolecular information processing. The system is designed to study the interplay of conformation, kinetics, structure, and dynamics in complex biochemical networks. As described in Part I of this two part paper optical signals may be used for tailoring initial conditions and harvesting the resulting output. Part II presents the architecture of the simulator and describes sample runs that indicate how it can be used to investigate structure changes in natural biological information processing and their potential significance for information technology applications.

Information processing operations of optical biomolecular devices based on nonlinear dynamic mechanisms are considered. For the case of pseudo 2D versions these devices can be described in terms of neural networks having lateral connections. Potentialities for further development of these devices are considered.

The methods of creating of systems which properties depend both on an outer physical signal and enzyme activity are discussed. The construction of conjugation chain between responses of stimuli-sensitive materials and urea hydrolysis catalyzed with urease is demonstrated on the basis of light- and thermoreversible polymeric matrices.

We present a model for color vision system with learning capabilities. The system adapts to statistical properties of its input. The adaptation is done by utilizing unsupervised learning techniques, as self-organizing feature maps and vectorial boundary adaptation maps. A color difference reflecting statistical properties of input to the system is defined. The model was tested by using color data with different statistics and two different sets of rhodopsin- based color sensors.

The conception of information processing based on cyclic processes in living systems is offered. The basis of them are trophic modes of cells. Photochemical processes initiated by light radiation in purple membranes reproduce one of the prophic cycle of the halobacteriaes. Realization of static and dynamic structure formation in bacteriorodopsin containing films using modulated light flow seems to make conditions for neuron like interaction for cyclic processes. Input and processing information in this case is `deformation' of interrelated cyclic processes and decision making are transient processes initiated in a organized neuron like medium.

The development of the research on 3D Molecular Electron Transfer Memory (METM) based on double and single VLSI dynamic memory chips with molecular storage medium BChl- BPhe-Qa-Qb is regarded. Their rate performance functional and technology characteristics are compared. The simulation of the electron transfer in the storage medium and dependence of the noise lever from character of the storage information and number of the molecular layers are described. The principle opto-electrical circuit and algorithm of the performance for METM are presented.

The problem of image pre-processing by neuron-like algorithms concerns development of systems and methods of image processing in parallel regime. The processing algorithms of grey-tone images to sets of simplified binary images by software of neuron-like filtration (based on the models of homogeneous neural networks) are considered in the paper and partially realized. The time of image processing is calculated. Carrying out of the required set of operations with image is possible according to the convolution function view. The neural network parameters (the type of coupling function and type of element response to the external influence) are selected in accordance with the solution which we want to receive. The program uses function similar to the Mask Convolution function. This function as any other functions basic for image processing are realized in the Intel Corporation `Recognition Primitives Library (RPL) for the Pentium Processor' with great efficiency. Library works under Windows NT. Performed calculations confirm that neuron-like model may be applied for real time image processing using RPL.

At this paper it was supposed that the epileptiformic activity of the cortex can be explained by the self- excitation regimes of the interrupted sensor data transformation mechanism. To describe signal transformation processes, we used a three-component mathematical model of ensembles of neuron-like elements. Activity structures in the chosen `basic' models were analyzed by a special research system developed in the interface of the Windows operation system. We considered examples of the evolution of spatial activity structures by extraction of simple features of the initial signal as an analog of normal functional modes of the cortex. We found the parameters of the system which ensure burst modes of epileptiformic activity in a three-component model of neural network.

The spectrophotonietric titration of the IXc-bilirubiii aqueous solution with the rat brain neuronal (synaptosomal) enthrane particles aqueous suspension (p1120o 7,2) is followed. by the transformation of the original free bilirubin aqueous solution absorptionspectram of visible light (single maximum at 440 nih and. slight shoulder in the region of 410-k20 nra) into membrane.-bound bilirubin aqueous suspension spectrum (two maxima in the region Qf '4-60 and 500 mm, respectively). As a result of the bilirubin molecules binding with membrane particles very weak origial fluorescence of the free bile pigment turns lfltO somemore fluorescence of the membrane bound. bilid.iene in the aqueous medium (the excitation and. the emission maxima are at 430 and. 520 nra ,respectively).

For main classes of molecular charge-transfer (CT)-complexes used for xerographic media, the regularities of the changing of the gains of photoprocesses are estimated. It is shown that for homological arrays of CT complex molecules having a constant CT distance, a parameters of photoprocesses (such as a optical CT degree, dipole moment, quantum gains of thermalized coulomb connected pairs and of free charge carrier photogeneration) are connected uniquely with a energy and space parameters. Established regularities allow to predict a unknown values of gains for known-structure CT complex molecules and to estimate a limit value of photoprocesses gains and photosensitivity of media based on CT complexes and photoprocesses mentioned above.

Data of psychodiagnostics of personal relations have been classified using neuron networks. Neuron networks were educated on model and empirical conceptions of psychological images of people groups. The results are discussed with the use of data on importance of minimum efficiency functions and magnitudes of antigradients by input signals. In the model conception, not typical objects are easily distinguished. Accentuations predominant in a group have been determined.

Gaussian-beam propagation in the bio-optical material bacteriorhodopsin is studied with the consideration of the material's intensity-dependent absorption and refractive index modulation. The beam focusing size, focusing position and their dependence on the incident beam parameters are simulated.