Study on the fiber-optic perimeter sensor signal processor based on neural network classifier

Presents a fiber-optic sensing alarm signal processing technology. It has great marketing demand because of the Optical-fiber sensor with high sensitivity, anti-electromagnetic interference, high corrosion resistance, etc. However, it is a problem about the false alarm to system. We use wavelet noise reduction technology and time-frequency domain features to construct the probabilistic neural network classifiers. The result shows it can largely reduce the false signals alarm.