Hello, I am using Tracking Tools to track a rigid body which I tape to the top of my hand to get a rough track of my hand position.
I am conducting various tests to see how fast my hand moves when I make various gestures, and I am experiencing unusual data specifically when I make a rapid "swipe" gesture with my hand.
I broadcast frame data from Tracking Tools and get position data of the Rigid Body through NatNetML for my XNA 'game' in C#.
I am not modifying the received data in any way, and I was simply logging the RigidBody X Y and Z position in a log file.
Here is what I notice:
During a fast "swipe" gesture, the position data jumps much more quickly than expected, and a large difference between only two data points is shown, which I find strange since it is highly improbable that my hand moved so fast (about 122mm in 16.66 ms). What is more worrying is that the previous data points show absolutely no change in position until the data point right before the 'jump'. Here are some images to illustrate this:
Graph Jump Data
Raw Jump Data
The Images show the X position of my hand in mm from some origin in front of me. The time is recorded on the horizontal axis in total elapsed milliseconds of my program, and each frame with position values comes about ever 1/100th of a second (16.66ms) so the data points are 16.66ms apart as the camera is recording 100frames/s.
Can anyone explain this behavior of the groups of 4 to 5 values in which TrackingTools returns the exact same positions? I just find it hard to believe that the entire swipe gesture from start to finish took place over just 2 data points in 16ms and I would like to know if this is something happening in Tracking Tools as my code simply outputs the received positions without any modification.
2 posts • Page 1 of 1
Do you have the original .tim .tra and .cal files from this shoot that we could review?