2022 Volume 3 Issue J2 Pages 563-571
In Japan, which is rich in mountainous terrain, natural disasters such as landslides often occur in mountainous areas. One of the obstacles to rapid emergency and recovery activities is the decision of whether or not to allow large emergency vehicles to enter narrow roads. If the width of narrow roads could be determined accurately and easily, appropriate countermeasures and rescue plans could be made, and rescue and recovery operations could be carried out more quickly than in the past. In this study, we propose an automatic road width measurement method using a monocular camera mounted on a vehicle for the passable area obtained from deep learning. We conducted measurements on an actual mountain road and were able to consistently achieve the target accuracy under limited conditions.
