Kinematics and localization of soft tissue objects.
Investigators: JP Dr. Pott/Prof. Verl, Prof. P. Xu
Within the project, approaches from the modelling of multi-body systems shall be used to discretise the deformation of tissue thus deriving a model with a largely reduced order. The modelling aims at approximating the large deformation with simplified models in order to capture the gross behaviour of the object. Based on these reduced models optical localization techniques shall be adapted to estimate both the configuration and the pose of the objects as well as pose of the kinematic. To cope with the expected large optimization and matching problems, a parallel computational architecture is considered making use of state-of-the-art multi-core processes and graphic adapters. These estimated results are coupled with measured sensor signals to find an optimal interaction strategy and position (e.g., gripping position of soft objects).