To a problem on components operation of a distributed system for image processing and analysis

The modern tasks of processing and analysis of large scale arrays of graphics information demand real-time operation and considerable computational resources. The ideology of developing such systems does not fit the increasing requirements and does not involve the possibilities of these computational resources. Classifications of such systems and their architectures are conducted. The advantages and disadvantages of existing solutions are detected. New goals and tasks are considered, and research directions in the field are pointed out. The alternative system structure for large distributed graphical information array processing, modes of implementation of data exchange processes between separate components of this system, and methods of integration with existing information storage and visualization tools are offered. The main component of the tendered system architecture is the intelligent computing core realizing the principle of an expert system. The feature of its operation is encompassed by using a self-learning mode during operation. It allows not only to improve the quality of automatic processing, and to use expert knowledge, but also to accumulate its own experience. Possible application areas of such a system are considered.