Wavelet transform filtering method of optical fiber Raman temperature sensor

The traditional wavelet filtering threshold de-noising method is widely used in the optical fiber Raman sensor temperature measurement, but the weakly spontaneous Raman scattering signal is easily submerged by optical noise and random noise. So the method has a low efficiency. In order to increase the efficiency of signal denoising in the process of temperature measurement, we proposed a new method combine default threshold and the given threshold based on the traditional method. The experimental data of 25Km fiber Raman sensor is filtered by the new method and the results show that the new threshold method can eliminate the noise in the signal, and also preserve the information of temperature variations.