Controlling a 3D printed arm using surface EMG - Honours Thesis
Вставка
- Опубліковано 21 бер 2014
- For my Engineering honors thesis in 2013 I programmed an open source 3D printable hand (the Inmoov) to take signals from my own arm and use them to execute useful gestures.
This video describes how I did that and what I achieved. For more information on the Inmoov go to: www.inmoov.fr/
Where I bought my parts:
core-electronics.com.au/store...
Just awesome! I hope Hubert was well behaved for other adventures throughout your thesis. Thanks for sharing!
Excellent work!
Nice work, good explanations.
InMoov Robot.
Thanks gael. I hope to do some more work and potentially look at using your new prosthetic hand design. Im very excited to print it.
VERY IMPRESSIVE WORK!!
How did you classified signal for various movement.
hello nice project i did something similar, i would like to know which feature extraction algorithms you used? Please.
Hi Erin, I am working to develop curriculum for a community STEM program and your work is fascinating. Would it be possible to discuss ways to implement your research into such a curriculum?
How can i find the machine that picksup the pad signals.
What brand, model is it?
I like it because it is multi chanel.
Since it can rees sensors. can you also send (Contract Muscles)?
Kind regards,
You should have given credit to Gael from InMoov robot, as you used his 3D print files.
Do you think this could work with someone that has very limited muscle movement, as a sort of exoskeleton?
This method is dependant on getting a good emg signal from the wearer. Obviously on an able bodied person this is easier to achieve. If there is some signal attainable from the wearer there may be amplification and processing methods to achieve your outcome. However if the signal can not be distinguished from background noise or interference emg control may not be the best options for some one with weak muscle control.
Kick ass
What muscles did you used?
How it reads the electricity from arm....I mean how does this thing works?
what signal processing technique did you use? did you rectify your signal first and then low pass filter it i.e. linear envelope?
Yea, defiantly! The initial steps included filtering base line drift, rectification and low pass filtering to achieve a nice clean signal envelope. Both infinite impulse response (IIR) and finite impulse response (FIR) filters where considered for this filtering. Ultimately a IIR filter was chosen as it has zero phase shift in the time domain so I was able to maintain the time mapping of activation and deactivation.
I would like to know more about what you have done in terms of time mapping of the activation and deactivation (have you performed machine learning to recognize movement of your hand or amplitude-based threshold (ON_OFF control strategy)). May I have your email address to discuss this in more details please?
I'am so happy to see this vedio, but this is 3 years ago ,I don't know whether you can see it. I'm a student, and I need a equipment to aquire the semg signal. but I have no ideal which equipment I should choice, could you tell me which equipment you choice? if you can tell me your address is better? thanks you very much!
«Myoelectric signal recognition using artificial neural networks in real time»
Adrian Del Boca, Florida International University
hi , can u give me ur email cause i wanna ask u about robot arm and how to control it by virtual work ??