Software work...

A project log for bioloop

a biofeedback wristband with galvanic skin response: giving insight into emotions

qquuiinnqquuiinn 08/17/2015 at 09:310 Comments

I don't understand computer science very well, but I have developed a rudimentary "algorithm" (if you can call it that) to interpret the information from the sensor. It scans for maximum/minimum readings, uses the value in the middle to establish a baseline, and gives the output value as a distance from this baseline. So far, things appear to be a little strange. Data seems to move up and down with no reason given, and I have not been able to get it to respond to induced stress very well. This is kind of a big deal, and I need to get the algorithm to work before adding all of the extras to make the firmware complete, which involves moving to the attiny sometime soon. Other code that I have used looks at the trendline of the data instead, perhaps that would be more useful? Some sources suggest looking for "SCRs", but my sensor is not nearly good enough to do that. Perhaps I need to see the variation in data, as one website I looked at saw a "cleaner graph" as a sign of reduced stress, and a noisier dataset a sign of more stress. Anyway, I need to get something working ASAP so I can meet the prize deadlines. The main concurn is just making sure bioloop will work well for everyone.