Scholarlink

<p>Land use and land cover (LULC) classification is a vital component in urban planning and sustainable city development, offering essential insights into resource management, environmental conservation, and urban growth strategies. Remote sensing (RS) techniques, coupled with deep learning (DL) algorithms, can significantly enhance the precision of urban land categorization, supporting sustainable urbanization efforts. In this research, we focus on utilizing the U-net + + deep learning model for detailed LULC classification in urban environments. The study leverages high-resolution satellite imagery from MBRSC to classify key urban features such as built-up areas, water bodies, vegetation, roads, and unpaved regions. The results show that the U-net + + model outperforms traditional methods, achieving an overall accuracy of 0.998 and an average precision of 0.993, making it an effective tool for urban resource management and planning. This research highlights the potential of advanced DL models in promoting sustainable infrastructure and urban resilience, contributing to the UN’s Sustainable Development Goals (SDGs), particularly SDG 11: Sustainable Cities and Communities.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Evaluation of U-Net + + architectures in high resolution image classification for urban planning

  • 期刊论文

  • 发表日期：

    2025年10月21日

查看全文
  • Alireza Sharifi,
  • Mohammad Mahdi Safari

摘要

Land use and land cover (LULC) classification is a vital component in urban planning and sustainable city development, offering essential insights into resource management, environmental conservation, and urban growth strategies. Remote sensing (RS) techniques, coupled with deep learning (DL) algorithms, can significantly enhance the precision of urban land categorization, supporting sustainable urbanization efforts. In this research, we focus on utilizing the U-net + + deep learning model for detailed LULC classification in urban environments. The study leverages high-resolution satellite imagery from MBRSC to classify key urban features such as built-up areas, water bodies, vegetation, roads, and unpaved regions. The results show that the U-net + + model outperforms traditional methods, achieving an overall accuracy of 0.998 and an average precision of 0.993, making it an effective tool for urban resource management and planning. This research highlights the potential of advanced DL models in promoting sustainable infrastructure and urban resilience, contributing to the UN’s Sustainable Development Goals (SDGs), particularly SDG 11: Sustainable Cities and Communities.