Home
OpenTrack
OT Tutorials
JSvg
To Do
Links

OpenTrack is intended to be a much larger project than it is now. Specifically it aims to include the following components:

• OpenTrack2d - This is very usable for tracking at present as attested to its use at the following institutions: Penn State Brandywine (Physics), West Chester University (Biomechanics/Kinesiology), Harford Community College (Kinesiology-based research project), and A.I. duPont Hospital for Children (Gait Laboratory for VO2 energetics analysis). A number of features still need to be added:
  • Addition of virtual markers to help visualize center of mass and the like.-DONE
  • Addition of vectors for velocities, momentums, accelerations, and forces
  • Ability to create videos containing overlaid markers, vectors, and links.-DONE
  • Ability to synch scrolling widget to both graphs and video
  • Reading of EMG files from common vendors
  • Upgraded VideoPoint file format support
  • Usability changes as needed
  • True platform independence via HTML5 Canvas and Video tags - Mostly DONE
• OpenTrack2d in Teaching - This project has been started courtesy of my work at Penn State. I should have some physics and kinesiology labs up shortly. - This has been implemented in version 0.5.10. The labs are not yet perfect, but give some examples of what is possible. Labs are accessible via OpenTrack in the Tutorials tab.
• OpenTrack3d - This is still in the conceptual phase, but the idea is to be able to drive two or more video cameras and implement the DLT algorithm to do 3D motion capture. This project will probably be written in QT or the like and may make use of spca drivers or code from the wxCam project. This has been a moving target over the years, and since there has been so much to do with the 2d project it has fallen by the wayside. A simple implementation may involve synching up two p2d files from OpenTrack2d (having calibrated the camera positions), and attempting to do time-space interpolation with the DLT. Check back in a year (summer 2010) to see if this actually goes anywhere. The current state of the project is thus:
  • Using Velleman's k8055 USB-DAQ ($43) we can strobe light rings in synch with two cameras. By strobing we eliminate motion blur even at the relatively low frame rate of 15-30 fps. The light rings have been prototyped using parts from RadioShack, and an acrylic cutout from BigBlueSaw.com and cost about $25 each to produce not including labor.
  • Using any cameras with the popular spca5XX chip and a modification to Michel Xhaard's spcaview library, asynchronous video capture which stores a time stamp for every frame can be captured from up to two cameras. With a faster computer, the code has been written so that up to four cameras are supported in this mode.
  • Centroiding of markers illuminated by the above light rings has been written.
  • A basic Direct Linear Transform algorithm code has been written using the Gnu Scientic Library (GSL).
• JSvg Graphics Library for Mozilla This graphics library is useful for creating Scalable Vector Graphics plots of scientific data and has the following modes: scatter plot, line plot, bar plot, histogram, stacked barplot, and stacked histogram modes. Multiplotting of data on one graph is supported. There is the need to support pie graphs, true. But more importantly, this was written when I had the very basest understanding of SVG. It should be possible to keep all the data in a graph and do zooming and the like.