I'm humbled to see the overwhelming support and utility this project has contributed to so many users over the past 6+ years. There was actually one point when I began to search and tally the number of thesis/dissertation/conferences/journals that have cited this project... I gave up in the hundreds.
But life goes on, I have moved on to new endeavors, and as you probably know, Thalmic Labs no longer exists to produce nor support Myo. Because of these factors, I will no longer be actively supporting MyoMex.
As I recall (loosely), there are two things you need to use this project in addition to the MATLAB/Simulink dependencies:
- Myo Connect
- Myo SDK for Windows
One user commented on MathWorks FEX that you may be able to find these using the wayback machine. This is what I was able to find. Unfortunately, it's proving difficult to locate the SDK download using this method.
- Myo Connect
However, if you use Google there may be better results.
One of the top results here is another repo where you can find the required Windows dependencies. Click here!
Access data from Thalmic Labs' Myo Gesture Control Armband in m-code!
Myo SDK MATLAB MEX Wrapper is also available on MathWorks File Exchange here.
Thalmic Labs' Myo Gesture Control Armband (myo.com) features an Inertial Measurement Unit (IMU) and 8 surface Electromyography sensors (sEMG) in addition to a nice Windows SDK that allows developers to obtain access to this data!
On the surface, this package contains a simplified m-code class, MyoMex, that enables MATLAB users to stream data from Myo at up to 50Hz (IMU) and 200Hz (EMG and meta data) with only 1 command!
mm = MyoMex(); % Upon construction, MyoMex starts accumulating streaming data in its myoData property
m = mm.myoData; % get MyoData object
% Data is now being pushed into log properties of m named,
% quat_log, gyro_log, accel_log, emg_log, etc.
% Data acquisition is non-blocking, too!
mm.delete(); % clean up
The IMU data includes estimated quaternion (orientation), three-axis gyroscope (angular velocity), and three-axis accelerometer (linear acceleration).
The sEMG data includes 8 raw data channels plus the output of Myo's built-in gesture detection.
Here are some of the things that you'll find in this package,
- README.txt - Step-by-step instructions for prerequisite configuration
- install_myo_mex() - installation tool
- build_myo_mex() - MEX-file build tool
- MyoMex_Quickstart - Quickstart guide script with example code and liberal comments
- MyoMexGUI_Monitor - GUI implementation of streaming data with visualization of all available data
The location of the files myo32.dll and myo64.dll must be indicated in the PATH
environment variable in Windows. The following error results from failure to perform this step of configuration.
English:
Error using MyoMex (line XXX)
MEX-file 'myo_mex' failed to initialize with error:
'Invalid MEX-file '<absolute path to MyoMex>\MyoMex\myo_mex\myo_mex.mexw64': Can't find specified module.'
Spanish:
Error using MyoMex (line XXX)
MEX-file 'myo_mex' failed to initialize with error:
'Invalid MEX-file '<absolute path to MyoMex>\MyoMex\myo_mex\myo_mex.mexw64': No se puede encontrar el módulo especificado.'
The solution for this problem is to add the location of the Myo SDK bin directory to the PATH
environment variable in Windows.
First, we must determine the absolute path of this directory by looking at two popular locations where you may have extracted the Myo SDK files.
The root of C:\
:
C:\myo-sdk-win-0.9.0\bin\
Your user folder (and your username is "AliceAndBob")
C:\Users\AliceAndBob\myo-sdk-win-0.9.0\bin\
Now you've determined the absolute path to the Myo SDK bin directory, you must append this path to the value of the PATH
environment variable in Windows. You can follow this tutorial to learn how to add this path to PATH
in your version of Windows. A special thanks goes out to C. Spiewak for providing this link in the comments here.