3D Sensors for Table Tennins

A table tennis ball is light, with all the weight in the skin. As a consequence spin is very important.

It is simple to locate a ball in 3 space with LIDAR, however to determine the speed and direction of any rotation you need surface landmarks (that will not alias) that are invisible to the eye.

Some IR absorbing pattern would work, as long as it is white in the visible spectrum.

I have seen some TT playing robots, but they did not seem to be spin aware, neither creating nor reading spin.

Suggest searches for this

Thank you Aurizon for making learn LIDAR. First time I came across this abbreviation.

http://www.ittf.com/ittf_science/SSCenter/docs/Boyer%20E-1-OK.pdf

http://www.theverge.com/2012/7/16/3161893/millisecond-motion-tracking-camera-ping-pong

ping pong ninja is my hero!

...and don't leave out machine vision software as a tool

Yes, a marked surface is easily tracked in real time. Human players track spin by analysis of the bat motion etc.

Baseball pitches are tracked using multiple (3) cameras. Perhaps a similar setup (smaller scale) could be used to track the ball.

http://www.sportvision.com/baseball/pitchfx%C2%AE (Sorry, link no longer available.)

Some type of reflector or reflectors on the paddles could be tracked with one or more cameras to determine the spin put on the ball by the players.

A baseball is rich in trackable features. A TT ball is a featureless sphere - which is why that tracking video marked the ball. For a human playable ball, to make a bot player against a person you would need to program the machine to read the many players cues a human uses to measure the spin applied by his opponent.

A human can not beat a bot playing with a marked ball because precise spin analysis would allow it to have an advantage.

In real TT play spin rules the play because the spin can steer the ball in flight and affect the bounce off the paddle where the ball gets traction and that motion must be dealt with

I realise this is not the sort of analysis your talking about, but, one thing that really upsets me about the data on different bats and rubbers, is that every retailer uses their own proprietary and subjective scores for speed, spin, and control.

It would be relatively easy to measure each different rubber for:

1.) Coefficient of restitution.

2.) Static friction.

3.) Sliding (dynamic) friction.

The coefficient of restitution would probably need to be measured over a range of speeds (some of which are probably higher than the terminal velocity of a TT ball), so there is quite a lot of "creative" work to be done.

I'm not too familiar with modern rubbers but "back in the day" most top (attacking) players used either Super Sriver or Mark V. It always seemed clear to me that Super Sriver had much higher dynamic friction, whereas Mark V had higher static friction. As a result Super Sriver was favoured by counter attackers, where Mark V was better for loop against back spin.

To do the friction measurements you would need to use a TT ball glued to the bottom of a "captive" weight.

I realise that modern rubbers have other qualities which are not as easy to quantify in simple physical measurement, like angle of throw, but, I reckon that any retail site which made some attempt to actually measure the above simple properties would make a killing.

As an aside, I work with several Indian guys who tell me that GKI make rubbers which are as good as any over here at a fraction of the price, but, it's impossible to get them shipped to this country.

Table Tennis has rules on foam, glue and rubber.

about rubber, foam etc

rules etc

Yes: I lived through the "arms race": the first very fast very spinny (high friction) rubbers; the anti-loop (very slow, dead, low friction rubbers); the first people to put an attacking rubber on one side and an anti-loop on the other, and, "twiddle" the bat whilst playing; the introduction of the rule that if you had different rubbers then they had to be different colours; people buying two sheets of, say, Mark V, and, putting one in the oven for an hour; with the final; ruling that all bats had to be black on one side and red on the other; grass (long pimples), and, then of course glueing.

But, in the post above, I'm just talking about someone measuring, and, publishing the quantifiable properties of existing manufacturers rubbers.

GKI is a big manufacturer of TT equipment in India, and, I believe they stick within the rules: I would be interested to hear the OPs son's opinion on their rubbers.

I thank everyone for your very useful contributions. I have passed the information to my son.