Adding three dimensions too a two dimensional plane.

Simply put we no longer need depth sensing cameras. I am currently independently building AI too translate sign language and will soon be giving my program out for free. Along with that I am creating a project on testing for hearing in infants. I realized early on that having an AI draw directly from video was causing technical limitations. For example my firs test with this AI was too sit on my bed (About four feet from my camera) and pinch. As I did that i soon realized that my simple program of-

if id == 0

X1 = Xlocation

if id == 17

X2 = Xlocation

Distance = X1-X2

if Distance >= 10

Print("it worked")

The premise of it is quite simple. All that my code was supposed too do was minus the X location from the two fingers and test if the fingers are close. If so it would print a message.

Sadly for my project we hit a road bump. The issue was that at further distances the pixels (And consequently the fingers) were recorded as closer dur too there depth from the camera. As such I compensated with finding two stationary points on the subject and tracking there decrease in X and Y as the subjects depth changes. Then after that I added the value of the ROT (Rate Of Change) back into the digits. This eliminated depth and allowed for me too test for sign language while eliminating the variable of depth which plagued my code.