|
Hi Axel, what a great way to get usable data from a rolling shutter camera!
Unfortunately the derivative algorithms in Tracker require positions at equal time intervals. But "Derivative" is a Java interface so if you or your students are Java coders you might be able to implement a new Derivative class that will do what you want. Let me know if you want more info on that. How, in Logger Pro, do you find the non-constant-timestep derivatives?
BTW it is certainly possible to define your data variable t_corr in Tracker. Look at the Data Builder for that. It just isn't possible, currently, to get the derivatives. Hope this is useful. Doug
Is there a way to use a different column as time stamp, and base the calculation of angle and angular velocity on that column? And could Tracker calculate derivatives of a quantity with non-uniform time steps?
Background: In one of our intro physics lab experiments, students are recording a video clip of a rotating disk. Tracing a point on the disk, they calculate angle vs. time and angular velocity vs. time. The webcam introduces a strong rolling shutter effect, which makes the (uncorrected) angular velocity appear to oscillate. However, this can be compensated by adjusting the time stamp of each video frame according to
t_corr = t - t_line * y
where t_line is the readout time for one line in the webcam image, and y is the vertical pixel coordinate. See here for how efficient this correction works: target="_blank">http://galaxy.phy.cmich.edu/~axel/PHY17x/Rolling_Shutter_Correction.mp4 >
At the moment, we use Logger Pro for the video analysis, which has, however, been discontinued, and we are looking into Tracker as a possible replacement. Is there any way this rolling shutter correction could be done within Tracker? Any suggestions would be highly appreciated!
Axel
|
