Transforming Sketches into Realistic Images - Leveraging Machine Learning and Image Processing for Enhanced Architectural Visualization

Aug 3, 2026

Upload by Ezra

A research article proposes a machine learning and image processing workflow to convert architectural sketches into visually coherent, realistic images. The approach uses the Stable Diffusion deep text-to-image model while adding image-processing algorithms to better interpret sketch inputs before generation. The study applies this method in architectural design contexts, aiming to support designers in producing accurate and compelling visual representations. Effectiveness is assessed through qualitative evaluation, demonstrating how the method bridges initial sketches and photorealistic renderings while contributing practical insights to the integration of AI and ML in architecture.

LanguageEnglish

CategoryTechnology

FormatPDF

Pages9

Reading Time23 min

AccessFree to download

Introduction

Stable Diffusion for sketch-to-image generation

Proposed image-processing workflow

A brief review of AI generated architecture

Generative models (GANs and diffusion models)

Stable Diffusion and its text-to-image focus

Sakarya University Journal of Science SAUJS ISSN 1301-4048 | e-ISSN 2147-835X | Period Bimonthly | Founded: 1997 | Publisher Sakarya University | [http://www.saujs.sakarya.edu.tr/](http://www.saujs.sakarya.edu.tr/) Title: Transforming Sketches into Realistic Images: Leveraging Machine Learning and Image Processing for Enhanced Architectural Visualization Authors: İlker KARADAĞ Recieved: 23.06.2023 Accepted: 7.08.2023 Article Type: Research Article Volume: 27 Issue: 6 Month: December Year: 2023 Pages: 1209-1216 How to cite İlker KARADAĞ; (2023), Transforming Sketches into Realistic Images: Leveraging Machine Learning and Image Processing for Enhanced Architectural Visualization. Sakarya University Journal of Science, 27(6), 1209-1216, DOI: 10.16984/saufenbilder.1319166 Access link [https://dergipark.org.tr/en/pub/saufenbilder/issue/80994/1319166](https://dergipark.org.tr/en/pub/saufenbilder/issue/80994/1319166) New submission to SAUJS [http://dergipark.gov.tr/journal/1115/submission/start](http://dergipark.gov.tr/journal/1115/submission/start) Sakarya University Journal of Science 27(6), 1209-1216, 2023 Transforming Sketches into Realistic Images: Leveraging Machine Learning and Image Processing for Enhanced Architectural Visualization İlker KARADAĞ *1 Absftract This article presents a novel approach for transforming architectural sketches into realistic images through the utilization of machine learning and image processing techniques. The proposed method leverages the Stable Diffusion model, a deep learning framework specifically designed for text-to-image generation. By integrating image processing algorithms into the workflow, the model gains a better understanding of the input sketches, resulting in visually coherent and meaningful output images. The study explores the application of the Stable Diffusion model in the context of architectural design, showcasing its potential to enhance the visualization process and support designers in generating accurate and compelling representations. The efficacy of the method is evaluated through qualitative assessment, demonstrating its effectiveness in bridging the gap between initial sketches and photorealistic renderings. This research contributes to the growing body of knowledge on the integration of machine learning and image processing in architecture, providing insights and practical implications for architects, design professionals and researchers in the field. Keywords: Architectural visualization, sketch-to-image transformation, machine learning, image processing, stable diffusion model. 1. INTRODUCTION The field of architecture is characterized by constant evolution and transformation, driven by the inherent complexity and originality of the design process [1] . Recent advancementsin technology, particularly in artificial intelligence (AI) and machine learning (ML), have opened up new possibilities in architecture, revolutionizing design processes and equipping designers with intelligent and efficient tools to achieve innovative and impactful outcomes [2, 3] . Recent digital approaches play a pivotal role in shaping contemporary architectural design and inspiring alternative solutions by encompassing transformative principles and philosophies [4] . This article aims to introduce a machine learning-based method called Stable Diffusion, which was introduced in 2022 as a deep learning framework for text-to-image generation. While its primary application is generating detailed images based on textual descriptions, it also holds potential for other tasks such as modifying or expanding the content of an image [5] . Notably, our approach incorporates image processing * Corresponding author: [ilker.karadag@cbu.edu.tr](ilker.karadag@cbu.edu.tr) (İ. KARADAĞ) 1 Manisa Celal Bayar University, Türkiye ORCID: [https://orcid.org/0000-0001-7534-2839](https://orcid.org/0000-0001-7534-2839) Content of this journal is licensed unde

FAQs

What method is used to transform architectural sketches into realistic images?

The study uses the Stable Diffusion text-to-image deep learning model, integrated with additional image-processing algorithms to interpret sketch inputs before generation.

How does image processing improve the sketch-to-image transformation?

Image-processing algorithms scan and prepare the sketches prior to transformation, helping the model better understand the input and produce more visually coherent outputs.

How is the proposed approach evaluated?

The method’s efficacy is assessed through qualitative evaluation, focusing on how well outputs bridge the gap between initial sketches and photorealistic renderings.

Aug 3, 2026

Upload by Ezra