Digital Image Processing for Study of Color Design of Residential Apartment Building

Regularities of color design of residential apartment building are investigated using intelligent technologies for analyzing digital images. Based on the analysis of the collected images, spectral colors and their combinations were found characteristic of every major Russian city. The research was carried out using the universal cognitive-analytical system “EIDOS-X ++”, using three statistical and seven information models based on the methods of information theory. After machine learning the EIDOS-X ++ system using digital images of buildings, which were marked with the names of cities, an intellectual analysis of digital data was carried out. The analysis revealed colors typical of the largest Russian cities. As the results, the spectral colors and their shades characteristic of Yekaterinburg, Chelyabinsk, Nizhny Novgorod, Samara and Kazan are given, as well as which combinations of them are most relevant for the studied cities. Having decided on the options for color combinations, it becomes possible to select the necessary amount of a particular color for the facade and develop design options for the main facades of houses under construction located in the cities under study. It is worth noting that in the creative process of searching for the color design of buildings, a person cannot be replaced. Therefore, the Eidos application created by the authors should be perceived exclusively as a recommendation system. The developed application can simplify and reduce the amount of work of architects, designers and related specialties.