Targets Classification Based on Multi-sensor Data Fusion and Supervised Learning for Surveillance Application

In surveillance application scenarios, like border security and area monitoring, potential targets to be detected may be either an unarmed person, a soldier carrying ferrous weapon or a vehicle. Detection is the first phase of a monitoring process, followed by the target classification phase and finally their tracking if required. This work focuses on classification step, where we introduce our classification approach not too resource-intensive, easy to implement and suitable for large scale environment. For that, we used probabilistic reasoning techniques to address multi sensing data correlation and take advantage of multi-sensor data fusion, then, based on adopted fusion architecture, we implemented our trained classification model in a fusion node, to make the classification more accurate.