Article for IEEE MLSP 2017 DATA CHALLENGE
Deblending of simultaneous-source seismic data via periodicity-coded nonnegative matrix factorization
Tzu-Hao Lin, Yu Tsao
Research Center for Information Technology Innovation, Academia Sinica
To increase the efficiency of seismic acquisition, one needs to break down the simultaneous-source seismic data into single source responses by a procedure called deblending. In this study, we employed the periodicity-coded non-negative matrix factorization (PC-NMF) to separate the primary and the secondary sources recorded in Petroleum Geo-Services dataset. Due to the application of random delay times between consecutive shots, the two sources displayed different patterns among various shots. By combining the response matrices of successive receivers, the PC-NMF can learn the basic components and group the basic components into two clusters according to the periodicity among shot index. Therefore, the deblending of simultaneous-source seismic data can be effectively achieved in an unsupervised manner.
The full paper can be accessed in IEEE DATA PORT.