Automatic method based on pso-optimized vision-transformer for gas detection in 2d seismic images

One of the geophysical techniques most frequently utilized in the oil and gas (O\&G) sector for hydrocarbon prospecting is seismic reflection. The seismic reflection technique is essential for an estimate the location and volume of gas accumulations in various onshore fields. However, this technique generates a large amount of data, and its data acquisitions are noisy. Thus it takes a while to analyze and interpret seismic data. Computational techniques based on machine learning have been proposed considering Direct Hydrocarbon Indicators (DHIs) to assist geoscientists in such activities. In this paper, we describe a method to detect gas accumulations based on the Particle Swarm Optimization (PSO) algorithm and the Vision Transformer neural network (ViT). In the best scenario, the proposed method achieved a sensitivity of 88.60%, a specificity of 99.56% and an accuracy of 99.37%. We present some tests performed on Parnaíba Basin and Netherlands F3-Block fields. Thus, it demonstrates that the proposed method is promising for assisting specialists in gas exploration tasks. DOI: 10.36558/rsc.v12i3.7902