| Autor: |
Utami, Wachyu Wiji, Slamin, Dafik, Agustin, Ika Hesti, Maylisa, Ika Nur, Baihaki, Rifki Ilham |
| Předmět: |
|
| Zdroj: |
AIP Conference Proceedings; 2024, Vol. 3176 Issue 1, p1-12, 12p |
| Abstrakt: |
Railway sleepers are a component of the railroad system that help to keep the rails stable. In rail sleepers, there are issues that are frequently faced, such as cracked and broken rail sleepers. If the damage is not discovered as soon as possible, this condition may worsen. Broken sleepers can put passengers at danger and lead into deadly accidents. Component aging and environmental factors that affect sleepers size and cause it to be imperfect can damage or displace the sleepers. Therefore, this research will examine the application of CNN (Convolutional Neural Network) to takes a preventive action on railway sleepers damage. This research utilize Alexnet deep learning architectures with three epochs, namely 50 epoch, 75 epoch, and 100 epoch. We will combine computer vision, deep learning, and drones to capture in this deep learning architecture. The research result shows that CNN performance gains 93.6667% accuracy, 100% precision, and 81% recall. It can be concluded that the CNN method using the Alexnet model with Epoch 100 and combined with Quadcopter drones can be used optimally to detect the damage of railway sleepers. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
