Classification of railway level crossing barrier and light signalling system using YOLOv3

Autor: Martin Kiac, Pavel Sikora, Malay Kishore Dutta
Rok vydání: 2020
Předmět:
Zdroj: TSP
DOI: 10.1109/tsp49548.2020.9163535
Popis: Nowadays, the world is experiencing an increasing boom in deep learning. This is more and more used in many areas such as medicine, robotics, industry, security systems, etc. This article deals with the detection and classification of railway barriers at level crossings, railway warnings, and light signaling systems. The evaluation system uses cameras, which are suitably positioned to capture the entire scene at a given railway level crossing. The detection itself is done using image processing techniques and deep neural networks. The proposed system uses the GPU acceleration to achieve real-time capability.
Databáze: OpenAIRE