Efficient under-sampled high resolution radon transform

Radon transform has been widely utilized for seismic data processing and is part of the seismic workflow for the last a few decades. It is robust, easier to compute, well established mathematical theory, and faster method to compute it. This paper suggests a new method for the parabolic Radon transform that results in an under-sampled high resolution Radon transform. The proposed method is based on a non-linear sampling technique, known as compressive sensing. Compressive sensing has emerged as a new sampling technique, which can be used to acquire exploration seismic data with less number of samples. The proposed method was applied for the attenuation of multiple reflections from the recorded seismic data which is one of the key issues in the field of seismic data processing. The method was tested on synthetic seismic data with multiple reflections. Furthermore, it was compared with existing methods for low as well as high Radon transform. From the simulation results, it was clear that proposed method, not only reduces the number of measurements but it also produces high resolution Radon transform with less computation time. Therefore, since the new method requires fewer samples and less computational time than existing techniques, it is believed that the proposed method is an appropriate new addition to the existing methods for efficient high resolution sparse Radon transform computation.