Low-Cost Deep Learning for Building Detection with Application to Informal Urban Planning
Open Access
- 9 January 2026
- journal article
- research article
- Published by MDPI AG in ISPRS International Journal of Geo-Information
- Vol. 15 (1) , 36
Abstract
This article studies the application of deep neural networks for automatic building detection in aerial RGB images. Special focus is put on accuracy robustness in both well-structured and poorly planned urban scenarios, which pose significant challenges due to occlusions, irregular building layouts, and limited contextual cues. The applied methodology considers several CNNs using only RBG images as input, and both validation and transfer capabilities are studied. U-Net-based models achieve the highest single-model accuracy, with an Intersection over Union () of 0.9101. A soft-voting ensemble of the best U-Net models further increases performance, reaching a best ensemble of 0.9665, improving over state-of-the-art building detection methods on standard benchmarks. The approach demonstrates strong generalization using only RGB imagery, supporting scalable, low-cost applications in urban planning and geospatial analysis.
Keywords
This publication has 45 references indexed in Scilit:
- So2Sat POP - A Curated Benchmark Data Set for Population Estimation from Space on a Continental ScaleScientific Data, 2022
- A review of building detection from very high resolution optical remote sensing imagesGIScience & Remote Sensing, 2022
- Building Extraction With Vision TransformerIEEE Transactions on Geoscience and Remote Sensing, 2022
- Image Segmentation Using Deep Learning: A SurveyIEEE Transactions on Pattern Analysis and Machine Intelligence, 2021
- Automatic detection of building typology using deep learning methods on street level imagesBuilding and Environment, 2020
- Learning to extract buildings from ultra-high-resolution drone images and noisy labelsInternational Journal of Remote Sensing, 2020
- Urban areas analysis using satellite image segmentation and deep neural networkE3S Web of Conferences, 2019
- Change detection techniques for remote sensing applications: a surveyEarth Science Informatics, 2019
- Fusion of Hyperspectral and LiDAR Data Using Discriminant Correlation Analysis for Land Cover ClassificationIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2018
- Fully Convolutional Networks for Multisource Building Extraction From an Open Aerial and Satellite Imagery Data SetIEEE Transactions on Geoscience and Remote Sensing, 2018