Advanced Sensor Fusion for Railway Security - A Hierarchical Graph-Based Approach

Abstract

Ensuring effective security surveillance in the railway sector presents a number of significant challenges, largely due to the extensive and dispersed nature of railway infrastructure. To effectively address threats such as vandalism, trespassing and sabotage, it is essential to implement continuous and comprehensive monitoring systems. Despite recent advances in technology, the issue of false alarms and information overload remains a significant challenge, undermining the efficiency of security operations. To overcome these challenges, recent studies have investigated the use of multi-modal fusion systems, which have proven effective in reducing false alarms. This paper introduces a Hierarchical Fusion Graph (HFG) as a generalised approach for interpreting a fusion system as a directed acyclic graph (DAG), enhancing the scalability and flexibility of sensor data fusion.