Efficient seismic denoising techniques using robust principal component analysis
Ningyu Sha, Ming Yan, Youzuo Lin · 2019
Robust Principal Component Analysis (RPCA) based seismic denoising approaches yield promising results in separating useful seismic events from noise. However, current RPCA-based methods suffer from expensive computational costs, which hinders their wide applications in seismic data denoising and preprocessing. In this work, we develop a cost effective denoising technique based on RPCA. Instead of solving for the clean data and noise simultaneously, we alternatively update them. This approach admits a large stepsize and increases the speed. In addition, we improve the model by incorporating a non convex term. To verify the effectiveness of our technique, we applied our denoising technique to both synthetic and field reflection seismic data. From the numerical results, we observe that our denoising methods not only produce comparable or better denoising results but also yield efficient computational cost. Through comparison to other RPCA-based denoising methods, our method is at least 4–10× faster. Presentation Date: Wednesday, September 18, 2019 Session Start Time: 9:20 AM Presentation Time: 11:00 AM Location: Poster Station 1 Presentation Type: Poster