Course Content

Human motion tracking (Motion capture) plays a critical role in a wide range of fields including Healt care services, Virtual and Augmented reality, entertainments, and sports. In Health-care services, th result of motion tracking is applied for diagnosis and treatment of patients. In entertainment, motion captur is used to animate a 3D character in movies and games while sports use it for injury prevention an improving performances. Different applications and situations had led to the development of variou motion tracking systems. Among all the motion capture techniques, camera and marker-based systems are the most prevailin technique which have high accuracy. By attaching a marker to a joint,regardless of the complexity of huma motion, a joint location is tracked by tracking the marker location using image processing technique.

However, due to certain limitations like the need for multiple high-resolution cameras, restriction to th studio-like environment, a huge amount of data, and high cost, the motivation to use these systems hav been diminished. Nowadays, due to the recent advances in MEMS sensors, low cost, small size, light weigh and low energy consumption of inertial sensors make them more attractive for the research on huma motion capture. However, these sensors have their own restrictions. Inertial sensors (IMUs) consist of three orthogonal gyroscopes, accelerometers, and magnetometers. I IMU-based systems, a sensor is attached to one body segment. By fusing sensory data, segment orientatio can be estimated. Based on the estimated orientation, together with the length of each segment and th arranging relationship between segments, the motion of the whole body can be obtained. These system have no line-of-sight requirements, and no emitters to install. Thus, IMU-based systems can be applie in a variety of applications where a studio-like environment is not necessary. In inertial systems, due to th integration of gyroscope signal over time, results lead to drift errors. Another drawback is that IMUs are no well-suited for determining absolute location. For accurate and drift free orientation estimation man fusion algorithms have been reported combining the signals from gyroscopes, accelerometers an magnetometers. The Kalman filter and optimization function have become the accepted basis fo the majority of orientation estimation algorithms. 

The experiment was successfully done and satisfactory results were obtained. The other proposed method which was tested for one body segment and compared to the first solution was to use gyroscopes along with accelerometers. Even though using a gyroscope would improve the results significantly, due to the high cost of gyroscopes and time limitations this method was not implemented. However, using gyroscopes are highly recommended for future design. The 3- D virtual LSM used in this project to validate how well the system tracks.