Optical Pattern recognition algorithms on neural-logic equivalental models and demonstration of their prospectiveness and possible implementations.

Historic information regarding the appearence and creation of fundamentals of algebro-logical apparatus-"equivalental algebra" for description of neuro-nets paradigms and algorithms is considered which is unification of theory of neuron nets, linear algebra and the most generalized neuro-biology extended for matrix case. A survay is given of "equivalental models" of neuron nets (NN) and associative memory is suggested new, modified matrix-tenzor neurological equivalental models (MTNLEMS) are offered with double adaptive -equivalental weighing (DAEW) for spatial -non-invariant recognition (SNIR) and space-invariant rrecognition (SIR) of 2-D images (patterns). It is shown, that MTNLEMs DAEW are the most generalized, they can describe the processes in NN both within the frames of known paradigms and within new "equivalental" paradigm of non-interation type, and the computing process in NN under using the offered MTNLEMs DAEW is reduced to two-step and multistep algorithms and step-by-step matrix-tenter procedures(for SNIR) and procedures of defining of space-dependent equivalental functions from two images (for SIR). Possible architectures of NN and MTNLEMs DAEW SNIR are discussed on base of matrix-tenzor equivalentors (MTE), represented by two optical matrix-tenter multiplicators. For realization of MTMLEMs DAEW SDR its arhitecture is offered as modification of known correlators and convolution operation systems. The results of modelling are given which confirm the possibility of increasing the NN capacity with such MTNLEMs at least up to 1,1 ammount of neurons! A successful recognition has been received of both large dimensional (from 100x100 up to 150x150 pixels) two level 2D images, and multilevel 2D-images (49x75x8bit) in the models DAEW SNIR (with the number of neurons from 3000 to 20000) and in models DAEW SIR.