Shadi M. Neuro-like method and means of images multiplication in computer network

Object of research - process of coding of photorealistic images with losses; aim of the research - improvement of image compression ratio as a resulting of application of neural networks of 2D Kohonen chart; methods of investigation and equipment used - analytical, experimental; theoretical and practical results - methods of image coding are suggested and studied, using neural networks of 2D Kohonen chart type, characterized by high level of adaptation due to introduction of additional learning stage, that provides the enhancement of compression ratio and high quality of the image restorted after image coding, the possibility of hardware implementation of suggested approaches based on optoelectronic element base has been shown, technical implementation of Kohonen chart applying optoelectronic neuron on bispin-device has been suggested; the degree of implementation - within the branch; spheres (branches) of implementation - computer networks and systems.