Tweet Share

• Send by email

  Send by email

  

  Accurate measurement for exposure rebar on communication tunnelsby deep learning

  Many tunnels for accommodating communication cables are reinforced concrete (RC) structures, and steel reinforcement (rebar) is exposed in the concrete surface due to aging. This paper proposes an automatic method to measure the length of the exposed rebar in communication tunnels using semantic segmentation. Metal racks for cables are placed at regular intervals in the tunnels. In this method, first, the area of the metal racks is detected using the semantic segmentation in photographed images. Then, image scale (mm/pixel) is calculated from the interval of the metal racks. Next, the area of the exposed rebar is detected. Finally, the length of the exposed rebar is estimated on the basis of the image resolution. As a result, we found that the estimated length of the exposed rebar was 0.72% error and had the highest accuracy in comparison with the measured length using a scale. We were able to estimate the size of the deterioration from arbitrary images in the communication tunnel.

  Recipient :

  Name ^* :

  E-mail address ^* :

  ^* Required fields

  Send   or  Cancel

• Print