[tog] [091labs-private] posture sensors

myr iammyr at email.com
Wed Apr 8 12:20:29 CEST 2015

Hi Gerry, Will and Gleb,

thanks a million for your help! I've checked out kinect because I
remembered that a couple of years ago, I saw a version of kinect to
apply directly on the hand and track finger movements. However, i could
not find it anywhere (maybe it's not yet in production).

that one would have been something to consider. while the usual kinect
or another external camera, would not be feasible because this solution
should track the position everywhere. it should be wearable. at list
ideally, this is how i was planning to deploy it. i've seen some
wearable devices to track movements so it should be possible to do what
i have in mind. however, they usually do not consider the "angle".
correct me if i'm wrong. maybe i'd only need to hack one of them to add
the angle detection. i'd love to know whether any of you has a more
precise idea than mine on how to achieve this :) or wants to help ;p

Thanks a lot!

On 08/04/2015 17:59, myr wrote:
> Hi Gerry,
> basically I would like to support someone I know, on the task of
> keeping a correct posture.
> to give you an idea on the size of this issue, imagine that when for
> instance one shoulder is lower than the other, this person can not
> write for the pain felt in the arm. So the issue is pretty big :p same
> with the hips that cause pain in the knees.
> usually these body parts are only very slightly incorrectly
> positioned. that's why i mentioned that the sensors should be able to
> detect "with some accuracy".
> measuring the angle could be useful for detecting the incorrect
> position of wrists, knees and maybe parts of the feet.
> I'm not too sure about motion but maybe could be useful for reporting
> purposes. and acceleration indeed, you're right, doesn't sound too
> useful at all.
> i'm not too sure kinect could be small enough to be applied to all the
> different body parts. would you suggest to try it?
> Thanks a million, Gerry ;)
> Myriam
> On 08/04/2015 17:48, gerryk wrote:
>> The sensors you are describing are good at detecting
>> motion/acceleration and relative angle/position, but not so good at
>> detecting absolute positioning in 3d space.
>> Really, you are looking at something like a Kinect with some joint
>> markers, like ping-pong balls.
>> I have done some playing with magnetometers and gyros for quadcopters
>> and head-tracking (for gaming), but as I said, these will only give
>> you relative positioning and motion, not absolute.
>> Could you let me know more about the tasks you have in mind?
>> On 8 April 2015 at 10:40, myr <iammyr at email.com
>> <mailto:iammyr at email.com>> wrote:
>>     Hi all,
>>     I would like to assemble a system that is able to sense the
>>     position of
>>     different parts of the body with some accuracy.
>>     I imagined it should rely on accelerometer, gyroscope and maybe also
>>     magnetometer, for this purpose.
>>     Would you suggest to use a different set of sensors?
>>     Would you know whether a kit exists that I could buy, which
>>     already does
>>     all this? In that case I may just focus on calibrating them together
>>     otherwise, could you suggest which ones to buy so that the
>>     overall size
>>     of the sensing mote is still practical (quite small)?
>>     Thank you so much for your help!
>>     Myriam
>>     P.S. when is the deadline for submitting to the dublin makers?
>>     couldn't
>>     find it on the website
>>     --
>>     Add some things to: http://wiki.091labs.org/
>> -- 
>> Contact me securely at secure at gerryk.com <mailto:secure at gerryk.com>
>> Download my public key from http://gerryk.com/secure
>> My BTC Wallet *1KEVnFpKrzHrgfmmRTUeBLFsWxMp1EWWY5*
>> -- 
>> Add some things to: http://wiki.091labs.org/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.tog.ie/pipermail/tog/attachments/20150408/feb1a21e/attachment.html>

More information about the tog mailing list