Combined person tracking and classification in a network of chemical sensors

Abstract

Transportation infrastructures play a crucial role in the way of life and the economic vitality of a modern society. Access points like stations, airports or harbors are among the most critical elements in these infrastructures because they offer the possibility to bring in hazardous materials that can be used for attacks against people and the transportation network itself. A timely recognition of such threats is essential and can be significantly supported by systems that monitor critical areas continuously and call the security personnel in case of anomalies. We are describing the concept and the realization of an indoor security assistance system for real-time decision support. The system is specifically designed for the surveillance of entrance areas in transportation access facilities and consists of multiple heterogeneous sensors: Chemical sensors detecting hazardous materials provide data for the classification of persons. But due to their limited spatio-temporal resolution, a single chemical sensor cannot localize a substance and assign it to a person. We compensate for this deficiency by fusing the output of multiple, distributed chemical sensors with kinematical data from laser range-scanners. Both tracking and fusion of tracks with chemical attributes can be processed within a single framework called Probabilistic Multi-Hypothesis Tracking (PMHT). An extension of PMHT for dealing with classification measurements (PMHT-c) already exists. We show how PMHT-c can be applied to assign chemical attributes to person tracks. This affords the localization of threats and a timely notification of the security personnel.