Saturday, December 11, 2010
DIY 10-finger flex sensor gloves for possible VR or video game control
I built a pair of flex sensor gloves for capturing the motion of all ten fingers. This system uses individual flex sensors made by Spectra Symbol and a National Instruments analog capture device to record the flex sensors' values.
Subscribe to:
Post Comments (Atom)
24 comments:
Great example of finding ways to prototype quickly. No duct tape required!
hi..do you need any programming for the sensor?
Anonymous, the National Instruments device includes LabView SignalExpress, which is a free datalogger program. It doesn't require any programming.
hi...can u specify d model name of d glove....thx a lot
In the video, I show the package, mention the brand and model -- Under Armor: The Yard II.
hi, can you put up a schematic of the electronics?
ball, I am a little busy at the moment, but my next youtube video (channel: bkraz333) will be about op amps, and I will discuss the circuit that I used for these flex sensors.
Ive had an idea for gloves like this for a while. If you play guitar, one hand uses the frets and one plucks the strings. So essentially if you have 5 positions for each finger on your left/right hand you could create a real 'air' guitar.
Also if you add one of those sensors to an elbow sleeve you could have a full range across all the frets by bending the arm to go up and down the frets.
The image I get of someone playing this is pretty entertaining but as far as learning to play an instrument, software is the limit.
Hope that someday I can see that idea matured, and good luck on your future projects.
Zach, that's a great application for the flex-sensor gloves. The videos of someone playing the "real" air guitar would be great. You should do it!
what happens on the end when you move your fingers really fast? The sensors don't seem to react fast enough...
Egbert, the NI DAQ program is sampling so fast the finger motion does not appear as a sinusoidal graph, but just vertically-moving horizontal lines. However, the "hold-off" time or intra-sample time is causing the graphs to be somewhat discontinuous. With proper sampling setup, the data from fast finger movements will be captured with no problem.
Hi Ben wanted to know actually about 4 values (sensitivity, resolution, noise, accuracy measures of the sensors) do you have the values for your or do you have them in a manual?
Also wanted to know if you can tell me what is the sampling rate for this device in order to capture it in real time?
Anonymous, check out the sensor datasheet.
https://www.jameco.com/Jameco/Products/ProdDS/150551.pdf
The sensors themselves are analog, and can be sampled at any rate you want. In this case, the NI USB-6008 is sampling at about 1KHz, if I remember correctly.
I have not exhaustively characterized these sensors for their repeatability, noise, etc.
Thank you very much Ben, but can I record or the flex sensor resistace readings as I want to plot them against the time and perform some other calculattion. I am intending to plot the values over a 60 seconds period of time. So I am wondering how many values for resistances will I be able to get over this period.
Anonymous, you would generally convert the sensor's resistance to a voltage with either an opamp or simple voltage divider, then sample the voltage with the NI DAQ. You can sample at any rate up to the DAQ's maximum. Since finger movements aren't particularly fast, you would probably want to low-pass filter the signal and sample at a fairly low rate (between 10Hz and 1KHz). NI SignalExpress is free software that would let you capture 60 seconds worth of data and display a graph. Search for the USB-6008.
Many thanks for your swift replies. I recall doing something like this before with LabVIEW. There is some option that lets you view resistance directly upon connecting it to the Analog channel. So no need for the voltage intermediate transformation, right? It is as if you are viewing the values of the resistance via an ohmeter but with the ability to log data and present them in a graphical manner against time..!!
You are like our god.I wanted to do a project Gesture Through Voice Communication by using glove with flex sensor.I will convert the glove signals into the voice signal. so can you plz give us circuit diagram and program of your project.My email id is isiddhesh@hotmail.com
Hi Ben,
I'm doing a project similar to what 'Anonymous' mentioned above called "Gesture Vocalizer" to convert hand gesture into voice signal.
So can you please send me the circuit diagram for using the flex sensor and interfacing details.
If possible and your time permits please provide me guidance for my project.
My mail ID : sivakuls2006@gmail.com>
Thanks in advance.
Cheers.
HARIHARAN
I do not have a circuit diagram or written instructions of how I built this. Unfortunately, I also do not have time to write such a diagram or instructions. Feel free to post questions if you get stuck with your project.
hi sir
I'm doing a project named "voice aided automatic sign language recognition system" to convert hand gesture into voice signal.Can you please tell me how to get more deflection digital values from flex sensors and from where i will get the glove you incorporated in your project.So please reply me urgently.
my email id: sanketdjoshi0790@gmail.com
Thanx in advance
Hello! My name is Khaled and I live in India. I saw ur video. And i'm working on a project called "Hand talk using Flex Sensor" for my final year engineering project. But I couldn't find a suitable "gloves" to fix my flex sensor. Do u think u can send me the gloves which u used here? Tell me how much will it cost.
Khaled, eBay dude!
http://www.ebay.com/itm/NEW-MEN-Under-Armour-Baseball-Batting-Glove-UA-The-YARD-II-Grip-Bat-HeatGear-MD-/290695636852?pt=LH_DefaultDomain_210&hash=item43aecf6774
hi i am doing a project similar to this glove . i need your guidance plz do reply on my id. tc l_k.111@rediffmail.com
what is the overall project cost??
Post a Comment