Capturing Human Hand Kinematics for Object Grasping and Manipulation
MetadataShow full item record
The aim of this thesis is to create a low-cost sensor equipped glove using commercially available components that can be used to obtain position, velocity and acceleration data for individual fingers of a hand within an optical motion capture environment. Tracking the full degrees of freedoms of the hand and finger motions without any hindrances is a challenging task in optical motion capture measurements. Attaching markers on every finger and hand joint makes motion capture systems troublesome due to practical problems such as blind spots and/or obtaining higher derivative motion constraints, such as velocities and accelerations. To alleviate this, we propose a method to capture the hand and finger kinematics with a reduced set of optical markers. Additionally inertial sensors are attached to the fingertips to obtain linear acceleration measurements. For optimal velocity estimation, a Kinematic Kalman Filter (KKF) is implemented and its result is compared to the time derivative of the Motion Capture System measurement. The higher derivative specifications are related to contact and curvature constraints between the fingers and the grasped object and are later used in formulating the synthesis task for the design of robotic fingers and hands. A preliminary prototype device has been developed to obtain position, velocity and acceleration information of each fingertip by incorporating multiple accelerometers into the basic design of reduced marker set.
Ghosh, Shramana (2013). Capturing Human Hand Kinematics for Object Grasping and Manipulation. Master's thesis, Texas A&M University. Available electronically from