Intelligent fusion processing of data perceived through IR sensors of multiple robots

This paper addresses the fusion processing techniques of multi-sensor data perceived through the IR sensors of military surveillance robots, which are positioned within a limited range with a lose distance between each of the robots. To combine the multi-sensor data from the distributed battlefield robots, we propose a set of fusion rules to formulate a combined prediction from the multi-source data. Through the use of these fusion rules, we make informed predictions of the type of target, which are represented by the degrees of reliability denoted in mathematical probabilities. We have implemented three fusion operators to compare the capabilities of their fusion processing, and have experimented them in simulated, uncertain battlefield environments. We have also applied fusion techniques to a battlefield simulator with multiple surveillance robots and various types of targets. The experimental results show that the fusion of multi-sensor data from military robots can be successfully tested in randomly generated military scenarios.