Road Layer Detection and Volume Calculation Using UAV Technologies and Artificial Intelligence

Abstract

In traditional road construction monitoring, the methods commonly involve manual measurements and visual inspections, which are not only time-consuming but also prone to errors. To solve these problems, this study introduces an innovative method utilizing UAVs to capture high-resolution images of construction sites and a YOLO object detection model to identify different road layers from captured images. The drone-captured pictures help create detailed 3D maps, also known as dense point clouds. These maps are then used to measure the volume of the layers via four software applications to identify the most effective tool for UAV-based road layer analysis. The model is trained to identify different road layers from the images, allowing for precise segmentation. This new way of using drones and advanced software changes how road construction projects are managed. This method greatly improves how quickly and accurately we can monitor road construction, making it a big step forward in managing road projects.