Real Time Building Crack Detection and Prevention Using Unmanned Aerial System

Abstract

Concrete often develops cracks, which can affect the appearance, structural integrity, and safety of a building, as well as the long-term reliability of the structure. The use of Unmanned Aerial Vehicles (UAV) for inspection involves two phases: developing a modern crack detection system using Unmanned Aerial Systems (UAS) and wireless data transfer, and utilizing advanced crack detection algorithms. The first phase involves designing and testing hardware, followed by designing algorithms and implementing software. UAS allows for quick and accurate capture of crack photos without causing damage to the surroundings. This eliminates the need for manual inspection, allowing for rapid deployment. Additionally, it incorporates wireless data transfer, enabling real-time identification through live video feed, which can be accessed on a smartphone. By using image processing techniques to analyze the captured photos and measure the crack diameters, a Faster Region-Based Convolutional Neural Network (Faster R-CNN) can accurately and efficiently record the data and confirm the system’s performance when applied to structures with cracks.