HI there,
let us assume you placed your cam orthogonal to your plane of motion, and that your subject will be running on that plane in your footage. Let us further assume your cam's lens distortion is negligible.
Using a bottle as shown in your pic is OK, as far as you are not aiming to achieve the highest precision possible when measuring your stride length. Consider the definition of stride length & do not confuse it with step length - former is defined as distance covered from touch-down to touch-down by the same leg, the latter as the distance covered from touch-down of one foot to the touch-down of the other. However, if you are about to key in the value in some wearable device, double-check the definition of the manufacturer.
Still, imho, the best would be for you to place a box of known height & width at the edge of the white line & take a snapshot before the actual running record.
Then, when you are up to do the analysis, you can select the calibration grid, drag & drop it's edges onto the edges of the box, right-klick, key in the dimensions & confirm calibration.
Save your calibration as .kva & load it onto your running records (if you have performed several of them).
Why the grid but not the line? Because if your bottle is not perfectly aligned to the center of the cam (which seems to be the case on the pic), you will have a perspective error affecting the distance measured. If you use a box at the line, the calibration grid will correct for that, so you can measure not only distance but also angles with reasonable precision.
I invite everyone to check this tutorials:
https://www.youtube.com/watch?v=upNY9cM … mp;index=2
Just check the whole section on 2D video analysis by BiomechanicsMMU.
Hope it helps.
Best,