Scholarlink

<p>Surface cracks in concrete are early indicators of structural deterioration that can compromise long-term durability and safety. This study introduces LEAF, a real-time, nondestructive image processing system for detecting and quantifying cracks in glass fiber-reinforced concrete (GFRC) subjected to compressive loading. M20 grade concrete cubes were reinforced with 1.5% glass fibers by cement weight and tested after 28&#xa0;days of curing period. High-resolution videos were recorded during compression testing under and later decomposed into 40,000 image frames. These frames were processed using LabVIEW with NI Vision Assistant to measure crack length, width and area. The image analysis pipeline included grayscale conversion, global thresholding, morphological filtering, and calibration using a physical scale for converting pixel values into real-world dimensions. Quantitative results show a clear progression of cracking severity across specimens, with crack lengths ranged from 18.72 to 230.57&#xa0;mm, widths varied from 0.35 to12.60&#xa0;mm, and total crack areas spanned from 42.42 to1309.49 mm<sup>2</sup>. The LEAF system achieved a calibration accuracy of <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(\pm\)</EquationSource> <EquationSource Format="MATHML"><math> <mo>±</mo> </math></EquationSource> </InlineEquation> 0.46%, a segmentation accuracy exceeding 93%, and an overall measurement uncertainty of <InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(\pm\)</EquationSource> <EquationSource Format="MATHML"><math> <mo>±</mo> </math></EquationSource> </InlineEquation> 5-6%. Strong statistical correlations between load and crack evolution (<i>r</i> = 0.85, <i>p</i> &lt; 0.01) confirm the reliability of the extracted metrics. The proposed framework offers a repeatable and scalable approach for digital crack assessment in concrete, suitable for laboratory diagnostics and future integration with predictive maintenance systems.</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

LEAF: A LabVIEW-Enabled Analysis Framework for Digital Crack Quantification in Glass Fiber-Reinforced Concrete

  • 期刊论文

  • 发表日期：

    2026年2月16日

查看全文
  • M. Kalai Selvi,
  • R. Manjula Devi,
  • S. Anandaraj

摘要

Surface cracks in concrete are early indicators of structural deterioration that can compromise long-term durability and safety. This study introduces LEAF, a real-time, nondestructive image processing system for detecting and quantifying cracks in glass fiber-reinforced concrete (GFRC) subjected to compressive loading. M20 grade concrete cubes were reinforced with 1.5% glass fibers by cement weight and tested after 28 days of curing period. High-resolution videos were recorded during compression testing under and later decomposed into 40,000 image frames. These frames were processed using LabVIEW with NI Vision Assistant to measure crack length, width and area. The image analysis pipeline included grayscale conversion, global thresholding, morphological filtering, and calibration using a physical scale for converting pixel values into real-world dimensions. Quantitative results show a clear progression of cracking severity across specimens, with crack lengths ranged from 18.72 to 230.57 mm, widths varied from 0.35 to12.60 mm, and total crack areas spanned from 42.42 to1309.49 mm2. The LEAF system achieved a calibration accuracy of \(\pm\) ± 0.46%, a segmentation accuracy exceeding 93%, and an overall measurement uncertainty of \(\pm\) ± 5-6%. Strong statistical correlations between load and crack evolution (r = 0.85, p < 0.01) confirm the reliability of the extracted metrics. The proposed framework offers a repeatable and scalable approach for digital crack assessment in concrete, suitable for laboratory diagnostics and future integration with predictive maintenance systems.