Measurement Data Processing Techniques

Raw acquired data must be processed to remove errors and extract useful information.

Technique	Purpose
Scaling and calibration	Convert ADC counts to engineering units, correct offset and gain
Digital filtering	Low-pass, high-pass, notch, moving average, to remove noise
Linearization	Look-up table or polynomial for non-linear sensors
Averaging	Reduces random noise by $\sqrt{N}$
Smoothing	Moving average, median filter
Peak and threshold detection	Finds events such as the R wave in ECG
Frequency analysis (FFT)	Finds frequency content of the signal
Statistical analysis	Mean, standard deviation, variance, correlation
Curve fitting / regression	Fits a mathematical model to the data
Data compression	Reduces storage size

Moving average

$y[n] = \frac{1}{N}\sum_{k=0}^{N-1} x[n-k]$

Exam tip

List 6 techniques with a one-line purpose each and include the FFT and averaging.