Scholarlink

<p>The growing demand for highway maintenance has underscored the need for accurate and efficient automated road crack detection, a critical technology for achieving cost-effective and high-quality road upkeep. Despite extensive research efforts globally, existing 2D and 3D detection technologies face significant challenges in handling complex and variable road conditions. Issues such as noise, shadows, water stains, and other interferences often lead to false positives and missed detections, limiting the reliability of current methods. To address these challenges, this paper proposes an innovative road crack detection approach that integrates 2D laser images with 3D laser information. By employing Dempster–Shafer evidence theory, a fusion decision model is developed to leverage the complementary strengths of 2D and 3D data while mitigating their individual limitations. This fusion-based method enhances detection accuracy and significantly reduces false detection rates. The proposed approach lays a solid foundation for advancing scientific, precise, and intelligent road maintenance. It holds significant research and application value for modern infrastructure management, offering a robust solution to improve reliability, efficiency, and scalability in road crack detection technologies.</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

Intelligent Asphalt Pavement Crack Detection with 2D and 3D Feature Fusion Using D–S Evidence Theory

  • 期刊论文

  • 发表日期：

    2025年5月26日

查看全文
  • Yiyang Zhou,
  • Zhiyuan Gu,
  • Lin Li,
  • Gang Shen,
  • Xiangfei Cheng,
  • Wenting Luo,
  • Haizhu Lu,
  • Chao Zhang,
  • Hong Gan

摘要

The growing demand for highway maintenance has underscored the need for accurate and efficient automated road crack detection, a critical technology for achieving cost-effective and high-quality road upkeep. Despite extensive research efforts globally, existing 2D and 3D detection technologies face significant challenges in handling complex and variable road conditions. Issues such as noise, shadows, water stains, and other interferences often lead to false positives and missed detections, limiting the reliability of current methods. To address these challenges, this paper proposes an innovative road crack detection approach that integrates 2D laser images with 3D laser information. By employing Dempster–Shafer evidence theory, a fusion decision model is developed to leverage the complementary strengths of 2D and 3D data while mitigating their individual limitations. This fusion-based method enhances detection accuracy and significantly reduces false detection rates. The proposed approach lays a solid foundation for advancing scientific, precise, and intelligent road maintenance. It holds significant research and application value for modern infrastructure management, offering a robust solution to improve reliability, efficiency, and scalability in road crack detection technologies.