Real-Time Tracking of Human Arm Movements
The goal of this project is develop a portable system designed to capture the users arm movements. This data can be fed into other software to analyze and predict the said arm movement. The mobile device should be small and battery powered, and be able to be kept on the body of the person using it. A wired connection to other computer equipment should not be a requirement, but can be a possibility for transferring previously collected data.
The sensors on the users arms will also be custom built to allow them to be as small and low power as possible. The ideal situation would be to have the sensors small enough to wear under the users shirt, so as to not obstruct daily life. The battery should also last long enough to not require a frequent charge.
In the modern world, there is a need for systems to collect multiple types of sensor data.There is a lack of systems on the market that provide a large amount of flexibility for a wide number of applications. Applications of such systems include rehabilitation, movement assist, and movement tracking. In many cases embedded software is designed to be application specific and does not provide users with a flexible architecture for their application design. Our solution provides an economically friendly and flexible hardware platform and a robust Real Time Operating System (RTOS) to provide rapid development of downstream applications. An example of said downstream application is to track human arm movements through the use of Inertial Measurement Units (IMUs). IMUs provide acceleration, rotation, and magnetic orientation data. Due to the non-ideal behavior of the sensors, compounding errors in the calculation will increase over time. Estimation of the error over time from a priori drifts can be used to mitigate large errors in calculations. One such way to deal with drift is through the use of both Kalman and Complementary Filters. A kinematic model of a human arm can be created using the data from the filters. This kinematic model can be used to provide a visual representation of the human arm movements in real time.