Technique for automatic marker recognition augmented reality for images of architectural plans

Abstract

The article describes a technique for recognizing image markers of design and estimate documentation of the developer's architectural plans using computer vision technology. Existing off-theshelf technologies and basic detection methods are being explored. The classification of methods according to the range of work is described. A recognition algorithm is proposed. The analysis of the Canny operator is carried out and its possible modification is considered to increase the efficiency of work when implemented in Python using OpenCV. Approbation of the developed software module is given.