2025 Volume 37 Issue 1 Pages 562-566
In Japan, road appurtenances are inspected with limited manpower and budget. In these inspections, outdoor workers go to the site to take images of the road appurtenances, while indoor workers confirm the locations of the images based on GPS data and organize them into a ledger. This process requires a great deal of labor, making efficiency improvements essential. Existing studies have proposed methods to manage images of road appurtenances using a three-dimensional model or QR codes to associate images with specific inspection sites. In this study, we propose a method to capture images of road appurtenances using pointing and calling gestures, while simultaneously recording the type of facility and verbalizing the damaged part, to streamline the association process.
