Current state of knowledge about data treatment techniques of position-time in the field of the motor system in biomechanics
Measurements made using an image - based motion analysis system are contaminated with noise generated during the recording and digitizing procedures. So the sampled signal can be considered as the sum of a quantity of true information associated to the physical phenomenon that is taking place and a quantity of information, that is nothing to do with this, representing the systematic and random noise induced by lens distortion, erroneous marker placement, calibration errors, skin and marker movements, digitizing errors, digitizer resolution, etc. Due to the nature of numerical differentiation, unless the random noise is reduced it may be amplified to such an order that the estimated derivative values may contain more error than signal. The main purpose of this study is the introduction of the reader to the most important techniques to treat de position -time data in order to reduce the noise and differentiate the displacement data. Besides, objective criteria are defined to select the most efficient fit data technique in Biomechanics.
KEY WORDS: data processing, curve fitting, Splines, digital filtering, Fourier Series, Kinematics