Foreign Body Intrusion Detection Algorithm for Track Surveillance Systems in Smart Cities

As modern science and technology advance, artificial intelligence technology has been widely applied. Among them, the image processing technology based on machine vision represents an essential part of artificial intelligence. Besides, its role in production and life is made increasingly important. A significant research challenge posed to image processing technology is foreign body intrusion detection algorithm. Given the issues with the conventional rail foreign body intrusion recognition systems, a novel and simple algorithm is proposed in this paper that is unified to detect foreign body intrusion from sensors of track surveillance systems. Not only is our method capable to detect object foreign bodies rapidly and accurately, and it is effective in reducing the volume of big data needing to be stored in the track surveillance system.