| CPC G01V 1/345 (2013.01) [G06T 7/11 (2017.01); G06T 7/73 (2017.01); G06T 2207/20081 (2013.01); G06T 2207/20101 (2013.01)] | 20 Claims |

|
1. A method, comprising:
receiving seismic training data comprising a plurality of images each including a plurality of traces;
predicting a geographic location of a feature in at least some of the plurality of traces based on a graphical location of a maximum of an amplitude peak therein, wherein the geographic location identifies the feature as comprising a sea floor or not comprising a sea floor;
applying a respective geographic label to each geographic location, wherein the geographic labels identify the geographic locations as comprising the sea floor or not comprising the sea floor;
classifying pixels of the plurality of images as representing the feature or not representing the feature, using a semantic segmentation model;
adjusting the geographic labels based on the classification of the pixels;
training, using the adjusted geographic labels and the seismic training data, a machine-learning model to identify the feature;
identifying the feature in a different seismic data set using the trained machine-learning model; and
modifying a drilling plan based on the identified features as determined by the trained machine-learning model.
|