Laser & Optoelectronics Progress, Volume. 56, Issue 7, 071503(2019)
Seismic Signal Blind Denoising Based on W-Weighted Nuclear Norm Minimization
Figures & Tables(9)
Fig. 1. Output signal-to-noise ratio versus noise offset. (a) Negative offset; (b) positive offset
Fig. 3. Marine prestack seismic signal and denoising effects of different algorithms. (a) Original signal; (b) with 5% noise; (c) DTCWT; (d) curvelet; (e) WNNM-true; (f) W-WNNM
Fig. 4. Denoising performances of different algorithms versus noise level in marine signal processing. (a) SNR; (b) MSE
Fig. 5. Landshot seismic signal and denoising effects of different algorithms. (a) Original signal; (b) with 5% noise; (c) DTCWT; (d) curvelet; (e) WNNM-true; (f) W-WNNM
Fig. 6. Denoising performances of different algorithms versus noise level in land signal processing. (a) SNR; (b) MSE
Fig. 7. Land poststack seismic signal and denoising effects of four algorithms. (a) Original signal; (b) with 5% noise; (c) DTCWT; (d) curvelet; (e) WNNM-true; (f) W-WNNM
Table 1. Comparison of denoising results corresponding to different noise offsets
View tableTable 1. Comparison of denoising results corresponding to different noise offsets
σ offset /% σ /%1 2 5 10 15 -30 36.4192 31.4788 25.2062 20.5495 17.7511 -25 36.6875 31.925 25.9789 21.6062 18.987 -20 36.928 32.3233 26.6716 22.5518 20.11 -15 37.1236 32.637 27.2142 23.3238 21.0682 -10 37.2642 32.844 27.5759 23.8636 21.7701 -5 37.3465 32.9429 27.7589 24.1101 22.0685 0 37.3714 32.9469 27.7855 24.0775 22.0095 5 37.3427 32.8781 27.6812 23.8494 21.7565 10 37.266 32.7603 27.475 23.5166 21.4456 15 37.1489 32.6089 27.2014 23.1526 21.1344 20 36.9978 32.431 26.8853 22.8009 20.8557 25 36.8203 32.2365 26.5491 22.4823 20.6023 30 36.6254 32.031 26.205 22.2063 20.3703
Table 2. Comparison of performance indexes of different denoising algorithms for marine prestack seismic signal
View tableTable 2. Comparison of performance indexes of different denoising algorithms for marine prestack seismic signal
Noise level Denoising algorithm P SNR/dBS NR/dBM SEIteration number W-WNNM 33.7067 28.4410 0.0004 8 WNNM-true 33.6938 28.4281 0.0004 6 σ =5%WNNM-false 29.9779 24.7122 0.0012 9 Curvelet 29.2879 24.0222 0.0012 - DTCWT 29.5756 24.3099 0.0011 - W-WNNM 30.2620 24.9963 0.0009 8 WNNM-true 30.2450 24.9793 0.0009 7 σ =10%WNNM-false 28.7818 23.5161 0.0014 9 Curvelet 25.2194 19.9537 0.0030 - DTCWT 25.6928 20.4271 0.0027 - W-WNNM 28.1667 22.9010 0.0015 8 WNNM-true 28.1234 22.8577 0.0015 8 σ =15%WNNM-false 27.1431 21.8774 0.0020 10 Curvelet 23.4587 18.1930 0.0045 - DTCWT 23.4705 18.2048 0.0045 - W-WNNM 26.5166 21.2509 0.0022 8 WNNM-true 26.4632 21.1975 0.0023 9 σ =20%WNNM-false 25.8859 20.6202 0.0026 10 Curvelet 22.1451 16.8794 0.0061 - DTCWT 21.9975 16.7318 0.0063 - W-WNNM 25.7057 20.4400 0.0027 12 WNNM-true 25.6398 20.3741 0.0027 10 σ =25%WNNM-false 23.5991 18.3334 0.0049 15 Curvelet 21.1127 15.8471 0.0077 - DTCWT 20.8750 15.6093 0.0082 -
Tools
Get Citation
Copy Citation Text
Zhenjie Feng, Weixue Han. Seismic Signal Blind Denoising Based on W-Weighted Nuclear Norm Minimization[J]. Laser & Optoelectronics Progress, 2019, 56(7): 071503
Paper Information
Category: Machine Vision
Received: Sep. 4, 2018
Accepted: Oct. 26, 2018
Published Online: Jul. 30, 2019
The Author Email: Zhenjie Feng (49909413@qq.com)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20