A Modified Algorithm for Wavelet Threshold Denoising to the Ultrasonic Signal of Optical Fiber Defect

In the ultrasonic testing to optical fiber fusion defect, because of the noise pollution, some weak bounding defect is difficult to detect. A modified threshold function based on the wavelet multi-resolution analysis threshold denoising method was proposed, and applied on defect signal denoising. The result of actual defect signal denosing showed that this method was of advantage to get less mean-squared error and high signal-to-noise ratio reconstruction signal therefore, which had excellent denoising result.