查看文章

Fatalities at grade crossings accounted for an average of 33% of all railroad industry fatalities occurring in the past 10 years. As road traffic increases and high-speed rail deployments become more common in the United States, the number of fatalities is expected to remain a concern. Railroads have tackled this challenge through a combination of engineering, education, and enforcement campaigns. One of these efforts has been the increased deployment of security cameras throughout railroad networks. These video sources allow for the collection of big data to better understand grade crossing violation behaviors. However, monitoring these video feeds and extracting useful information requires prohibitive amounts of manual labor. This research utilizes state-of-the-art vision-based artificial intelligence (AI) techniques to record, recognize, and understand railroad video data in real time. This system’s …