Converting scanned images of seismic reflection data into SEG-Y format

Archives across the world contain vast amounts of old or “vintage” seismic reflection data, which are largely inaccessible for geo-scientific research, due to the out-dated media on which they are stored. Despite the age of these data, they often have great potential to be of use in modern day research. It is often the case that seismic reflection data within these archives are only available as a processed stacked section, printed on paper or film. In this study, a method for the conversion (vectorization) of scanned images of stacked reflection seismic data to standard SEG-Y format is presented. The method addresses data displayed with a line denoting the waveform, where areas on one side of the baseline are shaded (i.e. wiggle trace, variable fill). The method provides an improvement on other published methods utilized within currently available academic software. Unlike previous studies, the method used to detect trace baselines and to detect and remove timelines on the seismic image is described in detail. Furthermore, a quantitative analysis of the performance of the method is presented, showing that an average trace-to-trace correlation coefficient of between 0.8 and 0.95 can be achieved for typical plotting styles. Finally, a case study where the method is applied to vectorize over 1700 km of land seismic data from the island of Gotland (Sweden) is presented.