Fundamentals and Methods of Machine and Deep Learning

13Crack Detection in Civil Structures Using Deep Learning

Bijimalla Shiva Vamshi Krishna¹, Rishiikeshwer B.S.¹, J. Sanjay Raju¹, N. Bharathi²*, C. Venkatasubramanian^1† and G.R. Brindha^1‡

¹SASTRA Deemed University, Thanjavur, Tamil Nadu, India

²SRM Institute of Science and Technology, Tamil Nadu, India

Abstract

The safety monitoring process of the structure of any civil engineering work is the most significant task. The continuous monitoring for any abnormal state of the structure is predicted and severe damages can be prevented. It also depends on the other environmental parameters like load, nature of seasonal parameter, and soil type, not only in civil engineering, but also other industries, which makes efficient use of the technology. In mechanical engineering, internal parts have to be monitored and set alarm to give attention to prevent the major damage to the system. Flight internal engines, brake system in cars, etc., are monitored. The manual approach completely relies on the person’s knowledge, experience, and skillset which obviously differs and always has the possibility of lacking objectivity in terms of quantitative analysis. The manual inspection can be replaced with automatic crack detection using ML and DL with computer vision. Currently, more powerful and fast image detection and recognition technologies are applied. The entire theme is all about providing the overview in brief and envisages the reader to analyze the crack detection using convolutional neural ...