Raster image digitization system for field data
AssigneeSchlumberger Technology Corporation
InventorsOmkar Anil Gune, Purnaprajna Raghavendra Mangsuli, Atul Laxman Katole
Life of this patent
Estimating the term…
Abstract
A method can include receiving raster images of logs, where the logs include plots, tracks and headers and where, compared to the plots and the tracks, the headers are underrepresented; processing the raster images using a trained machine learning model to generate segmentation masks for the logs for identification of the plots, the tracks and the headers; and generating digitized versions of the logs using the segmentation masks, where each of the digitized versions of the logs includes a digitized plot with an associated digitized header and an associated digitized track.
Claims20
A method comprising:
- receiving raster images of downhole assembly logs comprising well information, wherein the downhole assembly logs comprise plots, tracks and headers and wherein, compared to the plots and the tracks, the headers are underrepresented;
- processing the raster images using a trained machine learning model to generate segmentation masks for the downhole assembly logs for identification of the plots, the tracks and the headers;
- adjusting the trained machine learning model using labeled synthetic data, wherein the labeled synthetic data comprises synthetic downhole assembly logs, and the labeled synthetic data is generated by grouping headers to generate groups, labeling at least one header from at least one of the groups, and replicating the at least one header;
- and
- generating digitized versions of the downhole assembly logs, wherein each of the digitized versions of the downhole assembly logs comprises a digitized plot with an associated digitized header and an associated digitized track.
A system comprising:
- one or more processors;
- memory accessible to at least one of the one or more processors;
- processor-executable instructions stored in the memory and executable to instruct the system to:
- receive raster images of downhole assembly logs comprising well information, wherein the downhole assembly logs comprise plots, tracks and headers and wherein, compared to the plots and the tracks, the headers are underrepresented;
- process the raster images using a trained machine learning model to generate segmentation masks for the downhole assembly logs for identification of the plots, the tracks and the headers;
- adjust the trained machine learning model using labeled synthetic data, wherein the labeled synthetic data comprises synthetic downhole assembly logs, and the labeled synthetic data is generated by grouping headers to generate groups, labeling at least one header from at least one of the groups, and replicating the at least one header;
- and
- generate digitized versions of the downhole assembly logs, wherein each of the digitized versions of the downhole assembly logs comprises a digitized plot with an associated digitized header and an associated digitized track.
A non-transitory, computer readable medium comprising instructions that, when executed by processing circuitry, cause the processing circuitry to perform operations comprising:
- receiving raster images of downhole assembly logs comprising well information, wherein the downhole assembly logs comprise plots, tracks and headers and wherein, compared to the plots and the tracks, the headers are underrepresented;
- processing the raster images using a trained machine learning model to generate segmentation masks for the downhole assembly logs for identification of the plots, the tracks and the headers;
- adjusting the trained machine learning model using labeled synthetic data, wherein the labeled synthetic data comprises synthetic downhole assembly logs, and the labeled synthetic data is generated by grouping headers to generate groups, labeling at least one header from at least one of the groups, and replicating the at least one header;
- and
- generating digitized versions of the downhole assembly logs using the segmentation masks, wherein each of the digitized versions of the downhole assembly logs comprises a digitized plot with an associated digitized header and an associated digitized track.
Dependencies are inferred from the claim text.