Integrated Visual Sensing for Pose Estimation
DescriptionPose estimation with sensors on board is important for many applications such as autonomous robots in intelligent transportation or humanoid robot servants. Information on pose is the basis for behavior decision in such robots. Due to limitations in the sensing systems, existing methods in pose estimation are either too low in accuracy in the estimated pose or too high in computational cost to be useful for online applications. This project aims to set up a new sensing system by integrating inertial sensors with active visual sensors and to develop an efficient methodology for pose estimation. This new visual sensing system takes advantage of the large fields of view of the omnidirectional cameras and uses inertial sensors to enhance the performance of the vision system. The project's output will be an integrated visual sensing system with pose estimation methodology. The potential applications of the integrated sensing system will go far beyond those for robotics, to include applications for industrial automation, entertainment and service industry.
|Effective start/end date||1/04/09 → 31/03/11|