AUTOMATED SYSTEM FOR EVALUATING 2D-IMAGE COMPOSITIONAL CHARACTERISTICS: CONFIGURING THE MATHEMATICAL MODEL

Abstract

The paper is devoted to the development of an automated system for evaluating the compositional characteristics of a 2D image. It presents a special genetic algorithm designed to customize the mathematical model described in the previous paper by the authors - «Automated system for evaluating 2D-image compositional characteristics: mathematical model». The whole system is based on the research of R. Arnheim, the first paper of the cycle is devoted to the description of the general concept - «Automated system for evaluating 2D-image compositional characteristics: concept». A review of methods potentially suitable for solving the problem is given, the selected method is justified, the adaptation of the selected method to the specifics of a specific task is given. A mathematical model adapted to work with the existing mathematical model using the new method is presented. The structure of the training sample, the special aspects of data collection are described. Data analysis and sorting is performed using the developed genetic algorithm by code reuse; the choice of method is justified. The obtained results are analyzed, visualization of the compositional parameters of simple scenes is presented for different groups of respondents identified during data sorting and analysis. The similarity with the results obtained and demonstrated by R. Arnheim in his book without the use of information and automated methods are demonstrated. Results of the research are presented.