Automated generation of optimal thematic architectural layouts using image processing

Abdullah AlOmani, Khaled El-Rayes

Research output: Contribution to journalReview articlepeer-review


Thematic architectural layouts of many of the iconic buildings around the world were inspired by images from nature. This architectural design task is often complex and challenging as it requires the creation of optimal architectural layouts that resemble the selected image from nature while complying with all owner specified space allocation requirements. To support architects in this challenging task, this paper presents the development of a novel methodology to support architects and designers in automating the generation of thematic architectural layout designs that are inspired by images from nature. The methodology utilizes image processing to segment the nature-inspired image and analyze the segmented image to generate regions using region and boundary extraction algorithm. The methodology also integrates two novel optimization algorithms that are designed to optimize room assignments in the generated layout to maximize compliance with designer-specified adjacency requirements and optimize room areas to maximize the functional and operational performance of the architectural layout. An application example was analyzed to illustrate the use of the developed methodology and its novel capabilities in automating the generation of optimal thematic architectural layouts that preserve the image theme while complying with room adjacency and rooms area requirements.

Original languageEnglish (US)
Article number103255
JournalAutomation in Construction
StatePublished - Sep 2020
Externally publishedYes


  • Algorithms
  • Automated
  • Nature-inspired
  • Optimization
  • Regions
  • Segmentation
  • Thematic layout

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Civil and Structural Engineering
  • Building and Construction


Dive into the research topics of 'Automated generation of optimal thematic architectural layouts using image processing'. Together they form a unique fingerprint.

Cite this