Efficient Seismic Image Super-Resolution

Seismic field data suffer from Low Resolution (LR) image quality. This is due to the limited number of sensors used in a given seismic survey, as well as physical and technological limitations of the sensors. To obtain high-resolution (HR) images, either higher density acquisitions are carried out, or advanced sensors are used in the field. Both of which introduce financial, computational, and memory costs. Moreover, to improve the resolution of already acquired seismic field data, another survey would be needed, and this is a time-consuming high-cost approach. We propose the application of state-of-the-art Deep Learning based Convolutional Neural Network (CNN) models that are built for image super-resolution (SR) to increase the resolution of seismic images. Since seismic images interpretation benefits from image analysis algorithms that utilize Human Visual System (HVS) characteristics, objective as well as subjective image quality assessment will be carried out to evaluate the SR model used for seismic images.