Motion Capture Glove using Flex sensors, accelerometer and Arduino Due. The visualization of the captured data has been done using Blender Game Engine.
Flex Sensors gives a resistance varying linearly to the bend offered to it. Pressure based Flex sensors were attached to the glove, 2 sensors on each finger, such that the working portion of the sensor would overlap the joints of each finger. A two axis accelerometer was used to measure the tilting of the arm about the wrist.
Data acquisition and Processing :
Data from the sensors were acquired as analog values using an Arduino Due and transmitted to the computer using Serial Communication via USB. The Data is accepted by PySerial and given as input to modeling software named Blender. The posture of the arm is finally rendered using the Blender Game Engine.
- Remote control of robotic arms (esp. surgical arms)
- Simplification of algorithms for machine learning and/or programming of hands in humanoids
- Accurate data acquisition for animation industry
- Closed loop feedback for humanoids (if applied on hands of robots)
Future developments proposed :
Usage of sensors fabricated from optical fiber, with embedded arrays of Fiber Bragg Gratings (FBGs) for high precision data acquisition, addition of touch-sense capability for complete sensory achievement and in Development of Haptic arm