Havrylyshyn O. Іnformation technology of formation and forecasting of quality of digitization process of old books

The thesis is devoted to the development of models of information technology of formation and forecasting of quality of the digitization process of old books taking into account the systematic analysis of factors and the degree of their influence on the digitization process. The analysis of the technological process of digitization of old books has been carried out in the work. The factors and their priority for consideration in improving and optimizing the digitization process have been determined. A structural model of the digitization process of old books has been formed, which includes preparatory operations, the operation of obtaining a digital master copy and digital processing of copies. A graphical model has been formed as an oriented graph that identifies the relationships between factors. A multilevel model of the priority influence of factors on the digitization process of old books has been synthesized and optimized by the ranking method, where the factors of the process receive the following ranks. A simulation model “Restoration 1.0” for compensation of the missing part of the image on the digitized background has been developed by filling it with appropriate pixels as a means of forming the quality of the digitization process. A model of logical inference has been formed, the formation of a universal term-set of values and the corresponding linguistic terms which reflect qualitative characteristics of priority factors has been carried out. A fuzzy knowledge base has been developed, which reproduces the influence of factors on the digitization process depending on the quality level of linguistic terms. The membership functions of linguistic variables and fuzzy logic equations, which determine the relationship between the functions, have been constructed, and a complex indicator of the predicted quality of the digitization process of old books has been calculated according to the «center of mass» principle. Keywords: digitization of old books, information technology, factors, models matrices, fuzzy logic, linguistic variables, membership functions, forecasting, complex quality indicator.