Open Access Open Access  Restricted Access Subscription or Fee Access

Survey on Techniques to Segment Salt Bodies in Seismic Images

Amrutha Rose, T Rajasenbagam

Abstract


The areas in the earth which are rich in oil and natural gas also have huge deposit of salt below the surface. So, it is very important for the hydrocarbon industries to acquire the needed resources. The important challenge of seismic imaging is to detect subsurface salt structure which is required for the identification of hydrocarbon reservoirs and drill path planning. Unfortunately, the accurate identification and classification of large salt deposits is very difficult because of its non-linearity, and seismic imaging techniques often require professionals to interpret the salt bodies. In several fields, Convolutional neural networks (CNNs) have been successfully applied, and have got accurate results. Because of the less availability of labeled data, the performance of the CNN has been reduced.Seismic image analysis on a huge volume of data requires more labor and it is a time-consuming task if performed manually. In the past years, several efforts have been made to automate or semi-automate the process of identifying the salt bodies. In addition, it requires experts for identifying the right geological features. Machine-learning algorithms have also been used in the past which performed better than the existing methods. Recently, deep learning methods have been employed in this area, which give much better results compared to machine learning algorithms. So, the most important thing is to identify the accurate method for this segmentation problem. In this study, several deep learning models has been described so that it will be helpful to identify the perfect method which will lead to accurate result compared to the alreadyavailable methods. In this study, several network models have been described which will help to chose appropriate method for this segmentation problem.


Keywords


Machine learning, Deep learning, CNN, segmentation, seismic image

Full Text:

PDF

Refbacks

  • There are currently no refbacks.


This site has been shifted to https://stmcomputers.stmjournals.com/