Posture Visualization with Processing and Kinect

This is a repository of visualization of user posture and skeleton with Processing and Kinect, including a set of Processing skectches and a compainion website.

This is part of a class project for 15-821 Mobile and Pervasive computing at Carnegie Mellon University, in which we aim to provide real-time body posture information, detect bad spinal positions such as slouching, and couch senoirs to prevent osteoporosis as well as other chronic diseases. We collect spinal status data with Kinect and a BNO055 Adafruit 9-DOF absolute orientation sensor.

Setup

• Kinect (model 1517)
• Processing (2.2.1) along with following libraries:
  • SimpleOpenNI for kinect communication
  • v3ga for blob detection
  • HE-Mesh for graphic stuff
  • OpenGL for 3D stuff
  • jbox2d for physics-related stuff
  • Toxiclibs
• D3.js for website visualization

Directory

• physics_skeleton: processing sketches for real-time posture visualization
• website: compainion website for longitudinal posture visualization and analysis

Credits

• Creator: Qian Yang (🏠);
• Advisors: Asim Smailagic, Dan Siewiorek; 👏👏
• Collaborators: Kyuin Lee (💬), Raghu Mulukutla(💬);

References

• Kinect skeleton by Antoine Puel
• Kinect physics example by Amnon Owed, edited by Arindam Sen
• Wblut library example Creating a point cloud with ArrayList