Concrete Surface Crack Detection with Convolutional Neural Network
  • Author(s): Harsh Gupta ; Namita Goyal ; Vandana Choudhary
  • Paper ID: 1703956
  • Page: 199-203
  • Published Date: 26-12-2022
  • Published In: Iconic Research And Engineering Journals
  • Publisher: IRE Journals
  • e-ISSN: 2456-8880
  • Volume/Issue: Volume 6 Issue 6 December-2022
Abstract

This paper presents utilizing machine learning to detecting cracks on concrete surfaces. The purpose of this system is to help detect the cracks in concrete infrastructure, specifically in places like bridges and tunnels, where there are fewer human workers. Using deep learning algorithms, we were able to create a robust crack and occlusion detector that can look at images of different resolutions and photos taken under different light conditions. To determine the best outcome in each experiment, the model's accuracy was noted. The most effective experiment for the dataset utilized in this study produced a model with accuracy of 98.12%, demonstrating the promise of deep learning for concrete crack identification. The developed CNN is trained on images at a resolution of 227x227 pixels and, as a result, records with an accuracy of roughly 98%.

Keywords

Concrete surface crack detection, crack detection, deep learning, Image Processing, Convolutional Neural Network, CNN

Citations

IRE Journals:
Harsh Gupta , Namita Goyal , Vandana Choudhary "Concrete Surface Crack Detection with Convolutional Neural Network" Iconic Research And Engineering Journals Volume 6 Issue 6 2022 Page 199-203

IEEE:
Harsh Gupta , Namita Goyal , Vandana Choudhary "Concrete Surface Crack Detection with Convolutional Neural Network" Iconic Research And Engineering Journals, 6(6)