首页> 外文期刊>Optical fiber technology >Noise suppression for phase-sensitive optical time-domain reflectometer based on non-local means filtering
【24h】

Noise suppression for phase-sensitive optical time-domain reflectometer based on non-local means filtering

机译:Noise suppression for phase-sensitive optical time-domain reflectometer based on non-local means filtering

获取原文
获取原文并翻译 | 示例
           

摘要

Phase-sensitive optical time-domain reflectometer (Phi-OTDR) has been widely used in various fields because of its unique advantages for long-haul measurements. In this paper, the result of Phi-OTDR is treated as a two dimensional spatial-temporal image which can be filtered by the technique of non-local means (NLM). To improve the calculation speed of NLM, fast NLM based on integral image is introduced. The noise suppression effect of three intrusion cases processed by fast NLM is studied by both simulations and experiments, including single-position intrusion, double-position intrusion and riding along the optical cable. Both peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) in the area of image processing are used for quantifying the effects of NLM filtering. The noise suppression ratio (NSR) based on root-mean-square (RMS) is introduced and applied to evaluate the noise suppression of Phi-OTDR. After fast NLM processing, it can be obtained that the NSR could increase to more than 10 dB, the PSNR could increase by more than 12 dB and the SSIM also increase from 0.9361 to 0.9931. Parameters of both search window and smoothing control parameter are adjusted to investigate the regularities of above three metrics in three intrusion cases and the suitable NLM parameters with the optimal performance has been discussed in the field tests. Both simulations and field test results show that NSR, PSNR and SSIM increase to a constant level with the increase of search window. Both PSNR and SSIM increase at first and then decrease while NSR directly increases to a fixed level as the smoothing control parameter increases.

著录项

获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号