Inpainting borehole images using Generative Adversarial Networks
This is an incremental improvement for reservoir evaluation and geological interpretation in the oil and gas domain.
The paper tackles the problem of gap filling in borehole images from wireline microresistivity tools by proposing a GAN-based inpainting method, and the results show it effectively handles large-scale missing pixels and generates realistic completions.
In this paper, we propose a GAN-based approach for gap filling in borehole images created by wireline microresistivity imaging tools. The proposed method utilizes a generator, global discriminator, and local discriminator to inpaint the missing regions of the image. The generator is based on an auto-encoder architecture with skip-connections, and the loss function used is the Wasserstein GAN loss. Our experiments on a dataset of borehole images demonstrate that the proposed model can effectively deal with large-scale missing pixels and generate realistic completion results. This approach can improve the quantitative evaluation of reservoirs and provide an essential basis for interpreting geological phenomena and reservoir parameters.