Scholarlink

<p>Unauthorized construction (UC) poses numerous challenges to sustainable urban planning and regulation. The processes used to identify unauthorized construction through manually conducted surveys and inspections are inefficient and often expensive. The phrase “unauthorized construction” can be applied to describe any erection, modification, or improvement without securing the necessary permissions or licenses, which violates any sanctioned architectural designs, zoning codes, and local government regulations. Structures that fall under this category may even be considered illegal buildings. Moreover, these processes were not efficient enough to meet the rapid pace of urbanization in developing nations. The present systematic literature review (SLR) evaluates the new uses of artificial intelligence (AI) and deep learning (DL) methods for the purpose of detecting unauthorized construction through satellite and drone imagery. For the purpose of conducting a systematic literature review of the selected scientific articles in order to assess their critical quality, a rigorous selection procedure is adopted during the years between 2015 and 2025. Systematic classification of the articles is done based on the architecture used for the implementation of the DL method and the type of extracted features. This study identified the key challenges of the existing studies related to the applications of DL. The key challenges identified were that the techniques have not yet been developed on a large scale and are expensive. The paper investigates the techniques for improving the DL model's ability to learn and make decisions successfully. The paper establishes a conceptual framework for future studies on this theme. The applications of the techniques developed on the basis of this paper for the detection of unauthorized construction are used to define the objective of sustainable and intelligent development for the intelligent, sustainable, and legal development of the urban area. The paper has found the central role played by the applications of the techniques in the acquisition of the above-mentioned objectives and applied the objectives to the attainment of sustainable development goals (SDGs). To the investigators and policymakers who find it more convenient to use computationally efficient and ethical methods in the observation of unauthorized construction processes as they pertain to the topics of the current paper, this paper can be viewed as a foundational one.</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

Deep learning and geospatial AI techniques for unauthorized construction detection using satellite and UAV imagery: a systematic literature review

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年8月3日

查看全文
  • Monali R. Korde,
  • Anjali S. More,
  • Sucheta Kolekar,
  • Yashashree S. Mahale

摘要

Unauthorized construction (UC) poses numerous challenges to sustainable urban planning and regulation. The processes used to identify unauthorized construction through manually conducted surveys and inspections are inefficient and often expensive. The phrase “unauthorized construction” can be applied to describe any erection, modification, or improvement without securing the necessary permissions or licenses, which violates any sanctioned architectural designs, zoning codes, and local government regulations. Structures that fall under this category may even be considered illegal buildings. Moreover, these processes were not efficient enough to meet the rapid pace of urbanization in developing nations. The present systematic literature review (SLR) evaluates the new uses of artificial intelligence (AI) and deep learning (DL) methods for the purpose of detecting unauthorized construction through satellite and drone imagery. For the purpose of conducting a systematic literature review of the selected scientific articles in order to assess their critical quality, a rigorous selection procedure is adopted during the years between 2015 and 2025. Systematic classification of the articles is done based on the architecture used for the implementation of the DL method and the type of extracted features. This study identified the key challenges of the existing studies related to the applications of DL. The key challenges identified were that the techniques have not yet been developed on a large scale and are expensive. The paper investigates the techniques for improving the DL model's ability to learn and make decisions successfully. The paper establishes a conceptual framework for future studies on this theme. The applications of the techniques developed on the basis of this paper for the detection of unauthorized construction are used to define the objective of sustainable and intelligent development for the intelligent, sustainable, and legal development of the urban area. The paper has found the central role played by the applications of the techniques in the acquisition of the above-mentioned objectives and applied the objectives to the attainment of sustainable development goals (SDGs). To the investigators and policymakers who find it more convenient to use computationally efficient and ethical methods in the observation of unauthorized construction processes as they pertain to the topics of the current paper, this paper can be viewed as a foundational one.