Effect of Data Augmentation Method in Applied Science Data-Based Salt Area Estimation with U-Net

Year 2024, Volume: 5 Issue: 2, 70 - 86, 28.10.2024

Betül Ağaoğlu (cebe) , İman Askerzade , Gazi Erkan Bostancı , Tolga Medeni

https://doi.org/10.70562/tubid.1474999

Abstract

Oil and natural gas rank first as energy inputs worldwide. Other subsurface resources, such as salt, provide clues to obtaining these natural resources. Salt accumulation areas are subsurface resources used to locate oil and gas fields. Seismic images, which are geological data, provide information for locating underground resources. Manual interpretation of these images requires expert knowledge and experience. This time-consuming and laborious method is also limited by the fact that it cannot be replicated. Deep learning is a very successful method for image segmentation in recent years. Automating the detection of subsurface reserves in seismic images using artificial intelligence methods reduces time, cost and workload factors. In this study, we aim to identify salt areas using U-net architecture on the salt identification challenge shared by TGS (the world’s leading geoscience data company) Salt Identification Challenge on kaggle.com. In addition, the effect of data augmentation methods on the designed system is investigated. The data set used in the system consists of seismic images that are combined together for automatic detection of salt mass. The study aims to obtain the highest accuracy and the lowest error rate to detect salt areas from seismic images. As a result of the study, the IoU (Intersection over Union) value of the system designed without data augmentation method is 0.9390, while the IoU value of the system designed using data augmentation method is 0.9445.

Keywords

Deep learning, CNN, U-net, salt reserve, seismic data

U-Net ile Uygulamalı Bilimlerde Veriye Dayalı Tuz Alanı Tahmininde Veri Arttırma Yönteminin Etkisi

Abstract

Dünya genelinde enerji girdisi olarak petrol ve doğal gaz birinci sıradadır. Diğer yeraltı kaynakları, tuz gibi, bu doğal kaynakları elde etmede ipuçları sağlamaktadır. Tuz birikim alanları, petrol ve gaz sahalarının tespiti için kullanılan yeraltı kaynaklarıdır. Jeolojik veri olan sismik görüntüler, yeraltı kaynaklarının konumunu belirlemede bilgi vermektedir. Bu görüntülerin manuel yorumu uzman bilgisi ve deneyim gerektirmektedir. Bu zaman alıcı ve zahmetli yöntemin tekrarlanamaması da sınırlı bir durum yaratmaktadır. Derin öğrenme, son yıllarda görüntü segmentasyonunda çok başarılı bir yöntemdir. Yapay zeka yöntemlerini kullanarak sismik görüntülerde yeraltı rezervlerinin tespitini otomatikleştirmek, zaman, maliyet ve iş yükü faktörlerini azaltmaktadır. Bu çalışmada, TGS Salt Identification Challenge tarafından kaggle.com'da paylaşılan tuz tanımlama zorluğu üzerine U-net mimarisi kullanılarak tuz alanlarını belirlemek amaçlanmaktadır. Ayrıca tasarlanan sisteme veri çoğaltma yöntemlerinin etkisi araştırılmıştır. Oluşturulan sistemde kullanılan veri seti, tuz kütlesinin otomatik tespiti için bir araya getirilmiş sismik görüntülerden oluşmaktadır. Çalışmada, sismik görüntülerden tuz alanlarını tespit etmek için en yüksek doğruluk ve en düşük hata oranını elde etmek hedeflenmektedir. Çalışmanın sonucunda, veri çoğaltma yöntemi kullanılmadan tasarlanan sistemin IoU değeri 0.9390, veri çoğaltma yöntemi kullanılarak tasarlanan sistemin IoU değeri ise 0.9445 olarak belirlenmiştir.

Keywords

Derin öğrenme, CNN, U-net, tuz rezerv, sismik veri

References

• 1. Hubbert MK. Energy resources: a report to the Committee on Natural Resources of the National Academy of Sciences–National Research Council. Washington, DC: National Academy of Sciences-National Research Council; 1962. Report No.: PB-222401.
• 2. Economides MJ, Wood DA. The state of natural gas. J Nat Gas Sci Eng. 2009;1(1-2):1-13.
• 3. Özkan YZ, Akbaba MA. Örneklemeden rapor etmeye adım adım maden kaynak tahmini. Jeoloji Müh Derg. 2013;37(2):141-58.
• 4. Coombes J. The art and science of resource estimation: a practical guide for geologists and engineers. Coombes Capability; 2008.
• 5. Asjad A, Mohamed D. A new approach for salt dome detection using a 3D multidirectional edge detector. Appl Geophys. 2015;12(3):334-42.
• 6. Wu X, et al. FaultSeg3D: Using synthetic data sets to train an end-to-end convolutional neural network for 3D seismic fault segmentation. Geophysics. 2019;84(3):IM35-IM45.
• 7. Shafiq MA, et al. Detection of salt-dome boundary surfaces in migrated seismic volumes using gradient of textures. In: SEG International Exposition and Annual Meeting. SEG; 2015.
• 8. [Internet] Marine seismic surveys: what you need to know. 2024. Available from: https://energyproducers.au/fact_sheets/marine-seismic-surveys-what-you-need-to-know/
• 9. Di H, Wang Z, AlRegib G. Deep convolutional neural networks for seismic salt-body delineation. In: AAPG Annual Convention and Exhibition; 2018. Vol. 2018.
• 10. Shafiq MA, et al. Salsi: A new seismic attribute for salt dome detection. In: 2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE; 2016.
• 11. [Internet] Salt domes. 2024. Available from: https://geology.com/stories/13/salt-domes/#google_vignette
• 12. Babakhin Y, Sanakoyeu A, Kitamura H. Semi-supervised segmentation of salt bodies in seismic images using an ensemble of convolutional neural networks. In: Pattern Recognition: 41st DAGM German Conference, DAGM GCPR 2019, Dortmund, Germany, September 10–13, 2019, Proceedings 41. Springer International Publishing; 2019.
• 13. Karslı S. Son gelişmeler ışığında Türkiye’de kaya gazı. Journal of the Institute of Science & Technology. 2015;5(3).
• 14. Halpert A, Clapp RG. Salt body segmentation with dip and frequency attributes. Stanford Exploration Project. 2008;113:1-12.
• 15. Bodapati JD, Sajja RK, Naralasetti V. An efficient approach for semantic segmentation of salt domes in seismic images using improved UNET architecture. Journal of The Institution of Engineers (India): Series B. 2023;104(3):569-78.
• 16. Karchevskiy M, Ashrapov I, Kozinkin L. Automatic salt deposits segmentation: A deep learning approach. arXiv preprint arXiv:1812.01429. 2018.
• 17. Ozdemir C, Dogan Y, Kaya Y. A new local pooling approach for convolutional neural network: local binary pattern. Multimedia Tools and Applications. 2024;83(12):34137-51.
• 18. Ozdemir C. Classification of brain tumors from MR images using a new CNN architecture. Traitement du Signal. 2023;40(2).
• 19. Njima W, et al. Deep learning based data recovery for localization. IEEE Access. 2020;8:175741-52.
• 20. Badrinarayanan V, Kendall A, Cipolla R. SegNet: A deep convolutional encoder-decoder architecture for image segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2017;39(12):2481-95.
• 21. Zhu H, et al. Training a seismogram discriminator based on ResNet. IEEE Transactions on Geoscience and Remote Sensing. 2020;59(8):7076-85.
• 22. [Internet] TGS Salt Identification Challenge. Kaggle. 2024. Available from: https://kaggle.com/competitions/tgs-salt-identification-challenge
• 23. Özdemir C. Meme ultrason görüntülerinde kanser hücre segmentasyonu için yeni bir FCN modeli. Afyon Kocatepe Üniversitesi Fen ve Mühendislik Bilimleri Dergisi. 2023;23(5):1160-70.
• 24. Siddique N, et al. U-Net and its variants for medical image segmentation: A review of theory and applications. IEEE Access. 2021;9:82031-57.
• 25. Zhang H, et al. SaltISCG: Interactive salt segmentation method based on CNN and graph cut. IEEE Transactions on Geoscience and Remote Sensing. 2022;60:1-14.
• 26. Li H, et al. Deep learning-based model for automatic salt rock segmentation. Rock Mechanics and Rock Engineering. 2022;55(6):3735-47.
• 27. Chen X, et al. A stronger baseline for seismic facies classification with less data. IEEE Transactions on Geoscience and Remote Sensing. 2022;60:1-10.
• 28. Zhao Y, et al. Boundary U-Net: A segmentation method to improve salt bodies identification accuracy. In: Frontier Computing: Proceedings of FC 2020. Springer Singapore; 2021.
• 29. Ronneberger O, Fischer P, Brox T. U-Net: Convolutional networks for biomedical image segmentation. In: Medical Image Computing and Computer-Assisted Intervention–MICCAI 2015: 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part III. Springer International Publishing; 2015.
• 30. Fabijańska A. Segmentation of corneal endothelium images using a U-Net-based convolutional neural network. Artificial Intelligence in Medicine. 2018;88:1-13.
• 31. Du G, et al. Medical image segmentation based on U-Net: A review. Journal of Imaging Science & Technology. 2020;64(2).
• 32. Wang H, Miao F. Building extraction from remote sensing images using deep residual U-Net. European Journal of Remote Sensing. 2022;55(1):71-85.
• 33. Tran L-A, Le M-H. Robust U-Net-based road lane markings detection for autonomous driving. In: 2019 International Conference on System Science and Engineering (ICSSE). IEEE; 2019.
• 34. Fan W, et al. Stochastic reconstruction of geological reservoir models based on a concurrent multi-stage U-Net generative adversarial network. Computers & Geosciences. 2024;105562.
• 35. Zhou H, et al. Salt interpretation with U-SaltNet. In: SEG International Exposition and Annual Meeting; 2020. SEG.
• 36. Bochu RR, Buddha NK. Salt segment identification in seismic images of earth surface using deep learning techniques. In: 2023 Second International Conference on Electronics and Renewable Systems (ICEARS). IEEE; 2023.
• 37. Guo J, et al. A deep supervised edge optimization algorithm for salt body segmentation. IEEE Geoscience and Remote Sensing Letters. 2020;18(10):1746-50.
• 38. Chung Y, Lu W, Tian X. Data cleansing for salt dome dataset with noise robust network on segmentation task. IEEE Geoscience and Remote Sensing Letters. 2022;19:1-5.
• 39. HajNasser Y. MultiResU-Net: Neural network for salt bodies delineation and QC manual interpretation. In: Offshore Technology Conference; 2021. OTC.
• 40. Geng Z, et al. Semisupervised salt segmentation using mean teacher. Interpretation. 2022;10(3):SE21-SE29.
• 41. Saad OM, et al. Self-attention fully convolutional dense nets for automatic salt segmentation. IEEE Transactions on Neural Networks and Learning Systems. 2022.
• 42. Xu Z, et al. 3D Salt-HSM: Salt segmentation method based on hybrid semi-supervised and multi-task learning. IEEE Transactions on Geoscience and Remote Sensing. 2023.
• 43. Özdemir C. Avg-topk: A new pooling method for convolutional neural networks. Expert Systems with Applications. 2023;223:119892.
• 44. Dogan Y. A new global pooling method for deep neural networks: global average of top-K max-pooling. Traitement du Signal. 2023;40(2).
• 45. Shorten C, Khoshgoftaar TM. A survey on image data augmentation for deep learning. Journal of Big Data. 2019;6(1):1-48.
• 46. Ozdemir C, Dogan Y, Kaya Y. RGB-Angle-Wheel: A new data augmentation method for deep learning models. Knowledge-Based Systems. 2024;291:111615.
• 47. Civitarese D, et al. Semantic segmentation of seismic images. arXiv preprint arXiv:1905.04307. 2019.
• 48. ul Islam MS. Using deep learning based methods to classify salt bodies in seismic images. Journal of Applied Geophysics. 2020;178:104054.
• 49. Rezatofighi H, et al. Generalized intersection over union: A metric and a loss for bounding box regression. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition; 2019.
• 50. Bertels J, et al. Optimizing the Dice score and Jaccard index for medical image segmentation: Theory and practice. In: Medical Image Computing and Computer Assisted Intervention–MICCAI 2019: 22nd International Conference, Shenzhen, China, October 13–17, 2019, Proceedings, Part II. Springer International Publishing; 2019.
• 51. Eelbode T, et al. Optimization for medical image segmentation: Theory and practice when evaluating with Dice score or Jaccard index. IEEE Transactions on Medical Imaging. 2020;39(11):3679-90.
• 52. Guarido M, Li J, Cova R. Machine learning in geoscience: Using deep learning to solve the TGS Salt Identification challenge. CREWES Research Report. 2018;30:14.1–14.12.
• 53. Liu B, et al. Image segmentation of salt deposits using deep convolutional neural network. In: 2019 IEEE International Conference on Systems, Man and Cybernetics (SMC). IEEE; 2019.
• 54. Chung Y, Lu W, Tian X. Interactive segmentation using prior knowledge-based distance map. In: 2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS). IEEE; 2021.