Scholarlink

<p>Health monitoring of underwater concrete facility systems is important in civil engineering. Unlike conventional manual inspection techniques, digital image processing offers a more convenient and effective approach, becoming an indispensable tool for structural inspection. Cracks, which are pervasive defects, are a central focus of structural deterioration research. However, the complexity of the marine environment poses challenges to underwater visibility. In this study, the underwater environment under controlled laboratory conditions is replicated, where varying turbidity and illumination conditions and images of concrete cracks are captured. An approach combining a defogging algorithm with guided and fast guided filtering techniques is proposed to enhance both natural underwater images and crack images captured through experimental photography. When applied to turbid crack images captured under two different suspension conditions, the method increases the information entropy (IE) by 32.92% and 17.92% and the underwater color image quality evaluation (UCIQE) by 35.76% and 18.36%, respectively. These results demonstrate its efficiency in enhancing image definition. The findings of this study could significantly impact the practical applications of image visualization and evaluation for underwater concrete cracks.</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

Image Enhancement by Improving the Dark Channel Prior Method for Underwater Concrete Crack Detection

  • 期刊论文

  • 发表日期：

    2025年4月22日

查看全文
  • Yan-zhao Ma,
  • Bao-hua Chen,
  • Chen-xing Fan,
  • Qi-wei Lai,
  • Wu Zhao,
  • Zhe-xuan Yang,
  • Deng Li

摘要

Health monitoring of underwater concrete facility systems is important in civil engineering. Unlike conventional manual inspection techniques, digital image processing offers a more convenient and effective approach, becoming an indispensable tool for structural inspection. Cracks, which are pervasive defects, are a central focus of structural deterioration research. However, the complexity of the marine environment poses challenges to underwater visibility. In this study, the underwater environment under controlled laboratory conditions is replicated, where varying turbidity and illumination conditions and images of concrete cracks are captured. An approach combining a defogging algorithm with guided and fast guided filtering techniques is proposed to enhance both natural underwater images and crack images captured through experimental photography. When applied to turbid crack images captured under two different suspension conditions, the method increases the information entropy (IE) by 32.92% and 17.92% and the underwater color image quality evaluation (UCIQE) by 35.76% and 18.36%, respectively. These results demonstrate its efficiency in enhancing image definition. The findings of this study could significantly impact the practical applications of image visualization and evaluation for underwater concrete cracks.