Active perception for autonomous sensor systems

We have considered the problem of active perception interacting with a complex environment. The concept of perception provides a different means of awareness in an artificial system; it offers a classification process of the relevant features of an occurrence in the environment. We used a multidisciplinary perspective to show three active perception techniques: Direct perception, Sequential perception, and Direct perception with memory capabilities. The three case studies show that the way a system perceives the environment is highly dependent on the time sequence for the sensations and perceptions. Currently, we have developed the perception capabilities and the individual sensors in an artificial system. We are now in the process of fully integrating the perception concept from these three examples into a general platform.