Two-dimensional computer aided design drawings interpretation

Singh, Animesh

Permalink

https://hdl.handle.net/2142/130214 Copy

Description

Title

Two-dimensional computer aided design drawings interpretation

Author(s)

Singh, Animesh

Issue Date

2025-07-21

Director of Research (if dissertation) or Advisor (if thesis)

King, William Paul

Department of Study

Mechanical Sci & Engineering

Discipline

Mechanical Engineering

Degree Granting Institution

University of Illinois Urbana-Champaign

Degree Name

M.S.

Degree Level

Thesis

Keyword(s)

manufacturing, inspection, engineering drawing, CAD, algorithm

Abstract

The manufacturing industry continues to rely heavily on Two-Dimensional Computer Aided Design (2D CAD) drawings for inspection and quality control, despite the increasing adoption of Three-Dimensional Computer Aided Design (3D CAD) standards. The process of interpreting 2D CAD files is currently manual and relies on expert knowledge to analyze and extract dimensional and geometric information. The interpretation involves identifying key features, understanding annotation conventions, and mapping design specifications to physical components. The dependency of CAD Interpretation on human intelligence introduces inefficiencies and inconsistencies, limiting the scalability and automation of inspection processes. The representation of CAD files in a computer can be through a proprietary format like DWG (Drawing) or an open-source format like DXF (Drawing Exchange Format). The thesis describes an approach to automate the intelligent parsing of DXF file by extracting relevant geometric entities, associating dimensions with their respective features, and structuring the information into a JavaScript Object Notation (JSON) format for easy retrieval and processing. A critical enhancement to the representation of CAD files introduced in this research is the concept of "Critical Dimension CAD", which standardizes dimension placement, tolerance representation, and entity associations, thereby reducing inconsistencies in CAD drawings. The research methodology includes a two-stage algorithm: first, parsing and optimizing DXF data to eliminate redundant information, and second, applying heuristics to establish logical associations between dimensions and features. By automating the interpretation of 2D CAD drawings, this work reduces reliance on human expertise, enhances accuracy and consistency, and provides a foundation for integrating CAD interpretation with machine learning and computer vision techniques.

Graduation Semester

2025-08

Type of Resource

Thesis

Handle URL

https://hdl.handle.net/2142/130214

Copyright and License Information

Copyright 2025 Animesh Singh

Owning Collections