Automatic Construction and Visualization of 3D Architectural Models from 2D Floor Plan Images

Author

• Joey LederFollow

Date of Graduation

5-2026

Document Type

Thesis

Degree Name

Bachelor of Science in Computer Science

Degree Level

Undergraduate

Department

Computer Science and Computer Engineering

Advisor/Mentor

John M. Gauch

Committee Member

Susan Gauch

Second Committee Member

Alejandro Martin Gomez

Abstract

This research project focuses on Computer Vision and its capabilities through the medium of taking a floor plan as input and forming a cohesive 3D model out of it. The way this is accomplished is without any machine learning involvement, thus cutting out a significant storage and resource requirement that comes with it. However, it is incredibly difficult for a computer to distinguish features in an image because of how abstract an image is to a computer. That said, this thesis explores workarounds through filtering and other resources provided by libraries like OpenCV that are good for isolating everything that needs to be kept in the image. Though there isn't too much information being extracted from the image after filtering, this is a representation that with the correct filters and layering over an image we are capable of using a computer to isolate details from an image similarly to how machine learning is capable of digesting abstract concepts but with none of the drawbacks.

Keywords

Computer Vision; Floor Plans; Unity; Computer Graphics; Image Processing

Citation

Leder, J. (2026). Automatic Construction and Visualization of 3D Architectural Models from 2D Floor Plan Images. Electrical Engineering and Computer Science Undergraduate Honors Theses Retrieved from https://scholarworks.uark.edu/elcsuht/35