Machine Vision-Based Concrete Surface Quality Assessment

Manually inspecting concrete surface defects (e.g., cracks and air pockets) is not always reliable. Also, it is labor-intensive. In order to overcome these limitations, automated inspection using image processing techniques was proposed. However, the current work can only detect defects in an image without the ability to evaluate them. This paper presents a novel approach for automatically assessing the impact of 2 common surface defects (i.e., air pockets and discoloration). These 2 defects are first located using the developed detection methods. Their attributes, such as number of air pockets and area of discoloration regions, are then retrieved to calculate defects' visual impact ratios (VIRs). The appropriate threshold values for these VIRs are selected through a manual rating survey. This way, for a given concrete surface image, its quality in terms of air pockets and discoloration can be automatically measured by judging whether their VIRs are below the threshold values or not. The method presented in this paper was implemented in C++ and a database of concrete surface images was tested to validate its performance.

  • Availability:
  • Supplemental Notes:
    • Abstract reprinted with permission from ASCE
  • Authors:
    • Zhu, Zhenhua
    • Brilakis, Ioannis
  • Publication Date: 2010-2


  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01152173
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Mar 8 2010 5:33PM