Accurate upper body rehabilitation system using kinect.

Area of Science:

• Biomedical Engineering
• Rehabilitation Technology
• Human Motion Analysis

Background:

• Kinect offers a portable, low-cost, markerless solution for human motion capture in clinical settings.
• Kinect's accuracy in 3D joint localization often falls short of clinical standards compared to marker-based systems like Vicon.
• Variations in Kinect-derived body segment length and orientation limit its clinical utility.

Purpose of the Study:

• To develop and validate an optimization method for enhancing Kinect's accuracy in 3D human motion capture.
• To improve the precision of body segment length and orientation measurements using Kinect data.
• To enable more accurate clinical assessments and rehabilitation tracking using Kinect technology.

Main Methods:

• Proposed an optimization algorithm integrating Kinect Depth and RGB data.
• Implemented constraints for body segment length and orientation to refine joint center localization.
• Conducted an experimental study with ten healthy participants performing upper body range of motion exercises.

Main Results:

• Achieved a 72% reduction in body segment length variance.
• Demonstrated a 2° improvement in Range of Motion (ROM) angle accuracy.
• Validated the method's effectiveness for upper limb exercise measurements.

Conclusions:

• The proposed optimization method significantly enhances Kinect's accuracy for clinical motion analysis.
• This approach improves the reliability of measurements for rehabilitation and assessment applications.
• Kinect, with this optimization, shows greater potential for precise clinical use in upper limb evaluations.