ball tracking maths
Stephen's log entry: http://www.jugglingedge.com/log.php?LogDate=2017-03-14&UserID=913
[here's more editing space and eventually more maths gifted readers]
I'd go from there: https://en.wikipedia.org/wiki/Intercept_theorem, 1st graph (or better 3rd, below, ignoring "E" and "F" - "better" because ball's pictures' locations are parallel to camera's plane in the 3rd graph) .
So B and C then is the two camera's distance from each other. Where their common focus is, is S; its distance to the two cameras' plane can be measured from where the balls' pictures fully overlap. The ball's pictures then are A and B.
I just don't know how translate ball's pictures' distance in pixels as provided by tracking software into centimeters or inches in-real-distance, or, vice versa, how to translate the two camera's distance from one another in centimeters or inches into pixels. .. Holding two blue balls right onto the camera's lenses would make the software give the known cameras' distance also in pixels, but ball's pictures are then too big, so the blue tip of a needle might do it and still then I don't know, if cameras take any pics at all from object right on their lens.
But maybe using the known ball's diameter for gauging pixels versus inches or centimeters would work: find the distance at which the two ball's pictures touch. One then has a line segment AB in pixels for the ball's diameter's amount of centimeters or inches.
.. have to correct
So B and C then is the two camera's distance from each other. to "B and D".
The ball's pictures then are A and B. to "A and C".
Subscribe to this forum via RSS
1 article per branch
1 article per post
Green Eggs reports