COLOROMETRIC ANALYSIS OF THE IMAGE OF THE BUILDING SITE

Abstract

The paper considers the approach to the formalization of the colorimetric (color) atlas of the construction site (СABS), which is determined by the subject area at the intersection of disciplines: materials science, structural mechanics and machine mechanics, technical condition control, photogrammetry, pattern recognition theory. Each of the above applied areas contains its own knowledge base, various information technologies that take into account the characteristics of each of them. The procedure of forming the concept of color atlas of the construction site in the intellectual information environment in construction, its features and composition, the difference from the usual procedures of forming the knowledge system, as well as the proposed and described approach to its study, the information interaction of the intellectual environment of the object object of construction. The possibility of constructing different formalizations of СABS on the basis of ontological and semantic graphs and the graph of the construction site image is shown. The problems that arise in the process of formalization and analysis of СABS in order to search for image elements are identified and described. The substantive structure of СABS and the equivalence of СABS graphs are determined, taking into account the possibility of introducing similarity measures. СABS in terms of pragmatic classification is a specialized finding of elements on construction sites. The work is performed in the framework of cluster analysis and focuses on the development of methods for constructing algorithms for recognition and identification of construction sites. For the described transformations, research was conducted on their observation in digital images, as well as the possibility of observing the invariants of transformations. The work was performed for СABS in order to determine the possibility of building its information model, taking into account the development of methods for constructing algorithms for recognition and identification on construction sites. The possibility of constructing СABS on the basis of a set of images is investigated, the possibility of corresponding visualizations of СABS СABS as a means of categorization is established.