Sensor models and multisensor integration

Abstract

We maintain that the key to intelligent fusion of disparate sensory information is to provide an effective model of sensor capabilities. A sensor model is an abstraction of the actual sensing process. It describes the information a sensor is able to provide, how this information is limited by the environ ment, how it can be enhanced by information obtained from other sensors, and how it may be improved by active use of the physical sensing device. The importance of having a model of sensor performance is that capabilities can be esti mated a priori and, thus, sensor strategies developed in line with information requirements. We describe a technique for modeling sensors and the information they provide. This model treats each sensor as an individual decision maker, acting as a member of a team with common goals. Each sensor is considered as a source of uncertain geometric information, able to communicate to, and coordinate its activities with, other members of the sens ing team. We treat three components of this sensor model: the observation model, which describes a sensor's measure ment characteristics; the dependency model, which describes a sensor's dependence on information from other sources; and the state model, which describes how a sensor's observa tions are affected by its location and internal state. We show how this mechanism can be used to manipulate, communi cate, and integrate uncertain sensor observations. We show that these sensor models can deal effectively with cooperative, competitive, and complementary interactions between differ ent disparate information sources.