Road Network Deterioration Monitoring Using Aerial Images and Computer Vision
This addresses road maintenance challenges for local governments, but it is incremental as it builds on existing computer vision techniques for a specific application.
The authors tackled the problem of road network deterioration monitoring by developing a proof-of-concept protocol using aerial images and computer vision to automatically identify potholes and their severity, aiming to help local governments prioritize maintenance budgets and detect early deterioration to minimize costs.
Road maintenance is an essential process for guaranteeing the quality of transportation in any city. A crucial step towards effective road maintenance is the ability to update the inventory of the road network. We present a proof of concept of a protocol for maintaining said inventory based on the use of unmanned aerial vehicles to quickly collect images which are processed by a computer vision program that automatically identifies potholes and their severity. Our protocol aims to provide information to local governments to prioritise the road network maintenance budget, and to be able to detect early stages of road deterioration so as to minimise maintenance expenditure.