Shape Based Active Perception

Perception for artificial systems is increasingly gaining importance especially for systems with multiple sensors. For example, the autonomous vehicles equipped with camera, lidar, inertial measurement units, positioning systems etc are becoming popular. More importantly they are using the information from the multiple sources in the mean time, in the same coordinate system in order to gain information about the environment. Thus, the need for fusing the information from several sources in a common space for reasoning purposes is crucial. On the other hand, the position of the sensors from which the information content can be extracted efficiently depends also on the way that the base of the sensor is displaced in space and time. Thus, by having active perception sensor setup, the information acquired from the environment can be optimized by adopting an advanced planning algorithm for the vehicle displacement.

Start date (duration): 
November, 2015 to October, 2016
Project status and type: