Road damage is a common occurrence daily. Even in developed nations, road damage is still a possibility. Rainwater, temperature and weather variations, air temperature, construction materials on the road, subgrade conditions on the road, poor compaction process above the subgrade, and vehicle weights that exceed the limit are all factors that cause damage. Potholes are the most common form of road damage. Damaged roads are a major annoyance for drivers, can lead to accidents, and can even result in on-the-spot deaths due to falls since drivers are unaware of the potholes. If road damage is not discovered or overcome, it can be dangerous, and the road will deteriorate. Many media, including images, can be used to detect road damage. This study uses the image by utilizing the Faster R-CNN method to detect road damage. It reveals that using the MobilenetV2 backbone achieved the optimal performance indicated by the mAP value of 79.7%.
Detection Of Road Damage Using Faster Regional-Convolutional Neural Network Method
2022-11-22
657577 byte
Conference paper
Electronic Resource
English
Structural Damage Detection using Deep Convolutional Neural Network and Transfer Learning
Springer Verlag | 2019
|Structural Damage Detection using Deep Convolutional Neural Network and Transfer Learning
Online Contents | 2019
|