OBRAZ, INFORMACE, ENTROPIE.

Purpose – The text is a research study in the field of image analysis using the information entropy function and represents an alternative to the non-transparent work with images using neural networks, which is currently dominant. At the same time, it refers to an important moment of the intersection of science and art in the field of abstract art of the early 20th century. Design/Methodology/Approach – It describes a research in which the work of František Kupka was used as input image information in an experiment involving an original method of image analysis using the function of information entropy (Rényi entropy). Results – The design of the experiment included not only the computation of a function of image information entropy, but also human perception and their mutual comparison by regression analysis, and was designed as a pilot survey focused on the relationship between natural and simulated complexity. Originality/Value – At the same time, it points to the current state of the field of machine vision, its increasing opacity, and the need to better grasp it theoretically and practically. [ABSTRACT FROM AUTHOR]