THE POTENTIAL OF VARIOUS ARTIFICIAL INTELLIGENCE METHODS APPLICATION IN THE PROBLEM OF DRAWING RECOGNITION AND 2D 3D TRANSFORMATION
Open Access
- 29 September 2022
- journal article
- Published by Borys Grinchenko Kyiv University in Cybersecurity: Education, Science, Technique
- Vol. 1 (17), 21-30
- https://doi.org/10.28925/2663-4023.2022.17.2130
Abstract
The article analyzes the main methods of artificial intelligence in the task of recognizing drawings and transforming a 2D model into a 3D model. With the rapid development of information technologies, and especially in the pursuit of the most realistic reproduction of the project of the future product/house and other objects in digital form, the question of recognizing drawings and transforming a 2D model into a 3D model is very acute. As the number and complexity of tasks arising from the digitization of existing paper-based drawing and technical documentation grows, and the parallel need to transform two-dimensional models into three-dimensional models for visualization in three-dimensional space of complex objects, researchers have drawn attention to the possibilities of applying technologies and systems of artificial intelligence in the processes of drawing recognition and transformation of two-dimensional models into three-dimensional models. The first studies devoted to the application of artificial intelligence in the tasks of recognizing images on drawings began to appear in the early 90s of the 20th century. The analysis of approaches to the recognition of drawings allows us to consider the potential of using different methods of artificial intelligence in the task of recognizing drawings and transforming two-dimensional models into three-dimensional models. To analyze the potential of improving the work of CNN, as well as its architecture, without resorting to extensive expansion of the convolutional neural network (CNN) architecture, as well as taking into account the need to solve the task related to the logical vectorization of primitives and/or conditional graphics recognized by means of a convolutional neural network markings on drawings to perform 2D to 3D transformation. In the future, this stimulates researchers to look for alternative methods and models for image recognition systems on drawings.Keywords
This publication has 13 references indexed in Scilit:
- A Semantic-based Scene segmentation using convolutional neural networksAEU - International Journal of Electronics and Communications, 2020
- Using Artificial Intelligence Algorithms for Speech Therapy SystemsInternational Journal of Innovative Technology and Exploring Engineering, 2020
- Learning to sketch human facial portraits using personal styles by case-based reasoningMultimedia Tools and Applications, 2017
- A framework for similarity recognition of CAD modelsJournal of Computational Design and Engineering, 2016
- 3D geometry reconstruction from orthographic views: A method based on 3D image processing and data fittingComputers in Industry, 2013
- Using fuzzy logic to recognize geometric shapes interactivelyPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- Automatic learning and recognition of graphical symbols in engineering drawingsLecture Notes in Computer Science, 1996
- Document Analysis SystemsPublished by World Scientific Pub Co Pte Ltd ,1995
- Recognition of Hand-written Mechanical Drawing by Multi-Level Neural Network (MLNN)Published by Elsevier BV ,1994
- 3-D objects from 2-D orthographic views—A surveyComputers & Graphics, 1988