sophie4869/isure

Github Page: https://github.com/sophie4869/iSURE

Python: 2.7.6

MATLAB: R2014a

/HR: Heart rate detection

google-glass-sensors

This is the Google Glass project on Android Studio to record sensor data. It writes to sensordata0.txt on Glass, and can also post the data online. Change the 444 line of /google-glass-sensors/app/src/main/java/com/morkout/locationsensors/SensorActivity.java to the address of poSensorData.php file on your server.

rate.m

This is the MATLAB program that calculates the heart rate based on data file. From MATLAB, import the data and change the variable names to t, x, y and z.

HR_offline.py

Run this script to detect heart rate from data file. Change the data file path in line 14.

poSensorData.php

This posts sensor data to database. Change the database configuration to use. Don't forget to change the address of this php file in the Google Glass app.

HR_realtime.py

Change the database configuration in line 12 to use. This reads heart rate in real-time.

insert.py

A script to insert local data to the database. Change the configuration on line 5.

hello.py

A simple example of using SL4A.

/HAR: Human activity recognition

FeatureExtraction.py

This scripts extracts selected features from .csv data file. You can then convert the result to .arff format to use in Weka.

Notice: this script calculates statistic features(mean, variance, etc). Make sure that the input data file is of single label data. Combine different label data into a single .csv file AFTER the use of this script, and then convert to .arff file.

1. Copy this to the path where the train set and test set are
2. Run python path_to_data_set path_to_result.csv
3. Make sure the csv2arff class is in the same directory. Convert resulted .csv file to .arff file by: java csv2arff path_to_csv path_to_arff
4. Now you can use .arff file in Weka. Enjoy!