Automating Pre-Stack Migration Enhancement and Image Quality Analysis

Journal Title
Journal ISSN
Volume Title

The quality of a seismic migration image strongly depends on an accurate velocity model. However, the migration velocity model often contains errors causing artifacts that degrade seismic resolution and fidelity. Here, I propose a method to enhance the seismic migration image using processes in angle domain common image gathers (ADCIGs), without requiring modification to the velocity model. The ADCIGs impacted by velocity errors show non-flat reflections and scattered noise. To enhance the imaging quality in ADCIGs, these problems need to be treated before applying a stack. The ideal case of layers in ADCIGs should be flat without smear amplitude throughout all angles. Using this principle, I process one reflection at a time in each iteration by isolating it into a local window and then flattening it. Also, I apply internal processing steps in order to enhance the signal. The algorithm of segregating a reflection is called the connected-component labeling (CCL) method and is the leading technique to extract any feature that has a continuous form within the ADCIGs. Moreover, this method is insensitive to scattered noise so that it is a suitable approach to classify what is reflection or noise. To test the efficiency of this algorithm, I create ADCIGs which contain a relatively poor-quality image using an inaccurate migration velocity model. The primary objective is to show how well the CCL method can select various forms of reflections with associated noise levels. The results show that the CCL method is capable of decomposing a reflection individually, which can facilitate processing in the internal workflow. Finally, I evaluate the quality of processed migration images by comparing the processed Poynting-vector reverse time migration (PVRTM) images with reverse time migration (RTM) used Laplacian filter and partial stacked PVRTM. These comparable migration images are benchmarked against the synthetic reflector model being used as a baseline. As a result, the essential improvements of our method are addressed in the migration sections, namely the removal of diffraction artifacts at both small and large scales and the improvement of phase coherency. The final stacked migration image of the proposed method shows clearer geological features that are capable of delivering a more reliable seismic interpretation.

Poynting vector, Reverse time migration (RTM), Machine learning, SOMs, ADCIGs, Hough transform