Preprocessing is necessary in classification of CCPs for two major reasons. It is used to smooth out noises that
exist in them or extract useful features from CCPs’ signals. These features are supposed to draw out characteristics
of the original CCPs so they can be used as input vectors to the classifiers. Preprocessing adopted in previous
literature include heuristics [7], noise reduction or smoothing techniques [2] and image processing statistical features
[23] or combination of these techniques [18].