Abstract <p>In traditional alternating current field measurement (ACFM), branched cracks are typically treated as a single large crack during detection and repair. To ensure complete crack repair, it is often necessary to expand the repair area, resulting in a larger range of anti-corrosion coating polished. Therefore, it is crucial to introduce a new method that can accurately reconstruct the shape of branched cracks. This paper focuses on the key issues in reconstructing the shape of branched cracks and uses simulation and experimental methods to study the magnetic field distribution of these cracks. The magnetic field image generated by the crack is analyzed using image processing techniques, and the feature points of the image are extracted. The method of graph theory is introduced to research the reconstruction rules of branched cracks, and the method of automatic reconstruction for the surface shape of cracks is further studied. Additionally, this method is robust and achieves satisfactory reconstruction results even when missing data leads to low-resolution images. The experimental results show that the proposed method can achieve shape reconstruction of branched cracks with an error of 0.3–7.3% in branch crack length and 1.8–7.4% in branch crack angle. This method can reduce damage caused by excessive polish and provide the possibility for precise maintenance.</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

Surface Shape Reconstruction of Branched Cracks Based on Magnetic Field Image and Graph Theory Methods

  • Peng Gao,
  • Runze Ni,
  • Peipei Chen,
  • Linxi Wu,
  • Jiatao Guo,
  • Chuanhao Li

摘要

Abstract

In traditional alternating current field measurement (ACFM), branched cracks are typically treated as a single large crack during detection and repair. To ensure complete crack repair, it is often necessary to expand the repair area, resulting in a larger range of anti-corrosion coating polished. Therefore, it is crucial to introduce a new method that can accurately reconstruct the shape of branched cracks. This paper focuses on the key issues in reconstructing the shape of branched cracks and uses simulation and experimental methods to study the magnetic field distribution of these cracks. The magnetic field image generated by the crack is analyzed using image processing techniques, and the feature points of the image are extracted. The method of graph theory is introduced to research the reconstruction rules of branched cracks, and the method of automatic reconstruction for the surface shape of cracks is further studied. Additionally, this method is robust and achieves satisfactory reconstruction results even when missing data leads to low-resolution images. The experimental results show that the proposed method can achieve shape reconstruction of branched cracks with an error of 0.3–7.3% in branch crack length and 1.8–7.4% in branch crack angle. This method can reduce damage caused by excessive polish and provide the possibility for precise maintenance.