Automatic detection of characteristics tracks of optical signals received in multiparametric capillary and optical fiber sensors

The signals of time series is the beginning point of software signal processing in multiparameter optical capillary and fiber optic sensor. The time series signals have characteristic points and tracks. Moreover, despite of advanced electronic signal processing the digital signal is often modified with noise. Therefore, we perform analysis of selected algorithms for their use as automatic characteristic points and tracks detectors of the time series signals. For the experiment a wide range of algorithms was used. We implemented classical mathematic methods and algorithms known in medical apparatus as well as used in image processing. In particular, the paper presents results of evaluation of slope parameters detection and possibility of detection of characteristic saddle points of signal with wide range of allowed slope changes and significant noises which can cover partially the saddle track.

Author: Michal Borecki
Conference: Title