CAREER: Shape Localization, Recognition, and Reconstruction through Touch Sensing



The proposed research concerns computational principles of touch sensing in the context of part localization, model-based recognition, and shape reconstruction; it encompasses a number of areas of robotics - touch sensing, part orientation and localization, shape recognition, grasping, dextrous manipulation. The proposed localization scheme aims at determining the pose of an object with known geometry relative to the (robot) hand through contact. Algorithms will be developed for processing tactile data, computing geometric invariants, and registering them onto the object's surface. The education part of the project includes development of interdisciplinary courses with an emphasis on robotics and geometric computation.

2002-02-15 to 2008-01-31
Award Amount: 
Award Number: