Automatic events extraction in pre-stack seismic data based on edge detection in slant-stacked peak amplitude profiles

Zhao, Jing and Ren, Jinchang and Gao, Jinghuai and Tschannerl, Julius and Murray, Paul and Wang, Daxing (2019) Automatic events extraction in pre-stack seismic data based on edge detection in slant-stacked peak amplitude profiles. Journal of Petroleum Science and Engineering, 178. pp. 459-466. ISSN 0920-4105 (

[thumbnail of Zhao-etal-JPSE-2019-Automatic-events-extraction-in-pre-stack-seismic-data-based]
Text. Filename: Zhao_etal_JPSE_2019_Automatic_events_extraction_in_pre_stack_seismic_data_based.pdf
Accepted Author Manuscript
License: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 logo

Download (1MB)| Preview


Events picking is one of the fundamental tasks in interpreting seismic data. To extract the correct travel-time of reflected waves, picking events in a wide range of source-receiver offsets is needed. Compared to post-stack seismic data, pre-stack seismic data has an accurate horizon and abundant travel-time, amplitude, and frequency while the waveform of post-stack data is damaged by normal move-out (NMO) applications. In this paper, we focus on automatic event extraction from pre-stack reflection seismic data. With the deep development of oil-gas exploration, the difficulty of petroleum exploration is being increased. Auto recognition and picking of seismic horizon is presented as the basis for oil-gas detection. There is a correspondence between the real geology horizon and events of seismic profiles. As a result, firstly, recognizing and tracing continuous events from real seismic records are needed to acquire significant horizon locations. Picking events is in this context the recognition and tracing of waves reflected from the same interfaces according to kinematics and dynamic characteristics of seismic waves. Current extraction algorithms are well able to trace these events of the seismic profile and are undergoing great development and utilization. In this paper, a method is proposed to pick travel-time and local continuous events based on edges obtained by slant-stacked peak amplitude section (SSPA). How to calculate the SSPA section is discussed in detail. The new method can improve the efficiency and accuracy without windowing and manual picking of seed points. The event curves obtained from both the synthetic layered model and field record have validated the high accuracy and efficiency of the proposed methodology.