Deep Learning Advancements in Railway Track Segmentation: Previous Studies and Improvements

İngilizce (OpenAlex)

This article focuses on investigating the utilization of deep convolutional neural networks for segmenting railway tracks. Deep learning, which aims to simplify data processing by emulating human intelligence on computers, plays a significant role in this regard. Railway tracks are widely recognized for their importance in railway transportation. Consequently, ensuring track integrity requires thorough surface scanning. However, considering the extensive expanse of railway tracks, manual scanning proves to be a challenging and time-consuming task. Railway track segmentation serves as a fundamental step in identifying track defects, enabling easier detection by extracting tracks from surrounding images. This article discusses various studies conducted in this field and provides insights into the advantages offered by each approach.