A Deep Residual Network Designed for Detecting Cracks in Buildings of Historical Significance

This research paper investigates the application of deep learning techniques, specifically convolutional neural networks (CNNs), for crack detection in historical buildings. The study addresses the pressing need for non-invasive and efficient methods of assessing structural integrity in heritage conservation. Leveraging a dataset comprising images of historical building surfaces, the proposed CNN model demonstrates high accuracy and precision in identifying surface cracks. Through the integration of convolutional and fully connected layers, the model effectively distinguishes between positive and negative instances of cracks, facilitating automated detection processes. Visual representations of crack finding cases in ancient buildings validate the model's efficacy in real-world applications, offering tangible evidence of its capability to detect structural anomalies. While the study highlights the potential of deep learning algorithms in heritage preservation efforts, it also acknowledges challenges such as model generalization, computational complexity, and interpretability. Future research endeavors should focus on addressing these challenges and exploring new avenues for innovation to enhance the reliability and accessibility of crack detection technologies in cultural heritage conservation. Ultimately, this research contributes to the development of sustainable solutions for safeguarding architectural heritage, ensuring its preservation for future generations.

@article{Makhanova2024,
  title     = {A Deep Residual Network Designed for Detecting Cracks in Buildings of Historical Significance},
  journal   = {International Journal of Advanced Computer Science and Applications},
  volume    = {15},
  number    = {5},
  year      = {2024},
  publisher = {The Science and Information Organization},
  author    = {Zlikha Makhanova and Gulbakhram Beissenova and Almira Madiyarova and Marzhan Chazhabayeva and Gulsara Mambetaliyeva and Marzhan Suimenova and Guldana Shaimerdenova and Elmira Mussirepova and Aidos Baiburin},
  doi       = {10.14569/IJACSA.2024.0150558},
  url       = {https://doi.org/10.14569/IJACSA.2024.0150558}
}