Team Software Meeting 11.6.15

After Alex and Eric's meeting with Farhad, they worked at interpreting the data from the Fourier Transforms of their walking data using the specific time windows of five, ten, fifteen, and twenty seconds. From what they saw, there we three distinct peaks in the frequency graph, using our recorded step counts, they were able to see that the peak closest to zero corresponds to a double step count, the next peak would correspond to the single step count, and then the peak at 0 Hz, just corresponds to the mean of the accelerometer data being at a non-zero value. From there, Eric and Alex subtracted the means from the accelerometer data, in order to reduce the peak at 0 Hz. From there, we could tell that each accelerometer (x,y, and z), that the magnitude of these two peaks differed in each data set. The next step will be to determine how to work with these differences between each accelerometer.

Also, Alex looked into the current algorithms of other accelerometer functions in order to understand how to determine a step occured, which included normalizing the data and counting the amount of times it crossed zero, using the standard deviation of the data to establish thresholds that will be counted for steps, along with our procedure of using the highest frequency peak in order to determine the step frequency of the data.