Accelerated Fast BOTDA Assisted by Compressed Sensing and Image Denoising

We propose and experimentally demonstrate a scheme for accelerated fast BOTDA. The effect of signal-to-noise ratio (SNR) on recovery performance compressed sensing is simulated analyzed, it found that reduction in SNR requires much larger frequency data to recover the original Brillouin gain spectrum (BGS). To enable high probability, Block-Matching 3D filtering (BM3D) algorithm employed enhance time trace reduce number averages. Combining with principal component analysis (PCA) based technique, (BGS) can be successfully reconstructed from only 37.5% data. In experiment, 75 randomly selected acquired reconstruct BGS. Distributed strain achieved over 15 km single-mode fiber 3 m spatial resolution 0.52 MHz shift (BFS) uncertainty. Due process, measurement 40 averages less than 0.5 s.