AUTOMATIC CONSTRUCTION OF 2D-DRAWINGS FROM 3D-MODELS USING COMPUTER GRAPHICS TOOLS

Abstract

The article investigates the effectiveness of modern software for automating the creation of drawings from threedimensional models, which is a relevant task for the engineering and architectural industries. The main attention is paid to the analysis of the capabilities of AutoCAD, SolidWorks, Fusion 360 and algorithms based on convolutional neural networks (CNN). The study aims to determine the level of effectiveness of these tools in terms of key parameters such as processing speed, adaptation to complex shapes, and flexibility in customizing drawing files. The study showed that AutoCAD provides high speed processing of simple models, but has significant limitations in working with complex geometry, which reduces its effectiveness in tasks with a high level of detail. SolidWorks demonstrates better adaptation to curved shapes due to the use of spline algorithms, and also provides automatic updating of drawings when changing the three-dimensional model. Fusion 360 works effectively with organic shapes due to computational methods and cloud processing, but dependence on an Internet connection can slow down work with large amounts of data. CNN algorithms turned out to be the most effective for analyzing complex shapes, providing high detail, but require significant computing resources and time. It was concluded that the adaptation of software to complex shapes depends on the level of development of algorithms and the ability to flexibly configure drawings in accordance with standards such as ISO, ANSI or GOST. SolidWorks and Fusion 360 software products turned out to be the most balanced in terms of accuracy, adaptability and functionality. However, the development and implementation of neural networks open up new prospects for automation, especially in tasks with non-standard shapes and high complexity.