Thanks KP_PC but I have some questions more.
Let's think of a mass of 2 kg at the hand of the human.
He is doing such movements that can allow tha mass to be visible
to the camera.
In each frame, I detect where is the ball, so in a raw of
frame i can compute the vector of the velocity of the ball.
So I can compute the acceleration (a)
But what about the mass of the human arm?
I wonder this because I think it's logical that
the human muscles exert some force to raise the human arm
as well, apart from the ball.
What about it then?
And where does THIS mass (arm's mass) has it's center?
Of course I don't want to be fancy (spenting $),
so please inform me of the necessary details
of the experiment you described, and please
include any web pages or articles that might
this experiment has taken place with success.
"KP_PC" <k.p.collins at worldnet.att.net> wrote in message
news:ycaqb.206312$0v4.16283778 at bgtnsc04-news.ops.worldnet.att.net...
> "Panagiotis Artemiadis" <partem at mail.ntua.gr> wrote in message
> news:bob4n3$28qd$1 at ulysses.noc.ntua.gr...> | Do you have any idea of how can i measure
> | the position and joint torque of the human arm
> | during a movement?
> | Thank you in advance
> | P.A.
>> It's simple.
>> All that's required is Newton's f = ma.
>> Have your subject hold a standard mass in
> his hand.
>> Record video as he moves his arm.
>> Then, to your criterion for exactness,
> analyze the video. This can be done
> on a frame-by-frame basis, translating
> it all into 'levers', with respect to which,
> [since the mass of the arm is 'constant']
> the motion of the standard mass allows
> everything to be calculated.
>> Once you've collected a standardized
> set of motion-data [incorporating a wide
> range of accellerations ["range-widely :-]]
> you don't have to do any more data col-
> ection because you can translate it to an
> adequate(?) approximation of all pos-
> sible masses by just substituting numer-
> ical masses.
>> Or, if you have the $ to be 'fancy', put a
> little audio-transmitter in the mass, and
> disperse audio-receivers in a 3-D way
> that allows you to triangulate via doppler-
> shifts. [Polaroid Corporation used to sell
> everything needed [at reasonable rates],
> but that was a couple of decades ago, so
> I'm not sure if they still manufacture the
> kits.] Then you can write a computer
> program that will give you the information
> you seek in real 'time'.
>> Or, if you just want to have fun, 'suspend'
> your subject in a 'webbing' that's comprised
> of orthogonally-ordered spring scales. Here,
> the readings would be 'isometric', with the
> subject holding his position while you read
> all the scales. This last method would give
> a lot of smiles [and probably some deep
> insights into your subject's psychological
> predispositions :-]
>> k. p. collins