A new method for asynchronous multisensor information fusion

This paper presents a novel method for asynchronous multisensor information fusion (AMIF), which is real-time, model free and can provide the capability of reconfiguring sensors. The method is motivated by the idea that the information fusion is a procedure that the data of some sensors complement the data of others. The system computes the complements between the sensors. These complements are used to adjust the measurement at the moment of the fusion with a fuzzy inference system. The detailed principles axe first described and the algorithm is then implemented in a calligraphic robot, which includes three asynchronous sensors, image sensor, tactile sensor and sonar sensor.