US 12,461,262 B2
Classifying geologic features in seismic data through image analysis
Glenn K. Miers, Houston, TX (US); Dennis C. Furlaneto, The Woodlands, TX (US); and Brian D. Hughes, Houston, TX (US)
Assigned to ExxonMobil Technology and Engineering Company, Spring, TX (US)
Appl. No. 18/000,732
Filed by ExxonMobil Technology and Engineering Company, Spring, TX (US)
PCT Filed Jun. 9, 2021, PCT No. PCT/US2021/070678
§ 371(c)(1), (2) Date Dec. 5, 2022,
PCT Pub. No. WO2021/258096, PCT Pub. Date Dec. 23, 2021.
Claims priority of provisional application 62/705,282, filed on Jun. 19, 2020.
Prior Publication US 2023/0213671 A1, Jul. 6, 2023
Int. Cl. G01V 1/30 (2006.01); G06N 3/08 (2023.01); G01V 1/28 (2006.01)
CPC G01V 1/301 (2013.01) [G06N 3/08 (2013.01); G01V 1/288 (2013.01); G01V 2210/64 (2013.01)] 9 Claims
OG exemplary drawing
 
1. A method of identifying geologic formations within seismic data, comprising:
inputting a seismic data instance describing an underground region into a deep neural network that analyzes the seismic data instance at a 1:1 height to width ratio in a first network layer, a 2:1 height to width ratio in a second network layer, and a 4:1 height to width ratio in a third network layer to allow digestion of information from multiple aspect ratios simultaneously;
receiving from the deep neural network an indication of a geologic formation present in the seismic data instance; and
outputting the indication for display through a user interface.