Multimodality mineralogy segmentation system and method
US10107769B2 · kind B2 · utility
Assignee
Inventors
Key dates
| Filing date | Jan 11, 2016 |
| Grant date | Oct 23, 2018 |
| Priority date | — |
| Expiry date | Jan 11, 2036 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06V2201/06
- WIPO fieldMeasurement
- WIPO sectorInstruments
Abstract
A multimodality imaging system and method for mineralogy segmentation is disclosed. Image datasets of the sample are generated for one or more modalities, including x-ray and focused ion beam scanning electron microscope (FIB-SEM) modalities. Mineral maps are then created using Energy Dispersive X-ray spectroscopy (EDX) from at least part of the sample covered by the image datasets. The EDX mineral maps are applied as a mask to the image datasets to identify and label regions of minerals within the sample. Feature vectors are then extracted from the labeled regions via feature generators such as Gabor filters. Finally, machine learning training and classification algorithms such as Random Forest are applied to the extracted feature vectors to construct a segmented image representation of the sample that classifies the minerals within the sample.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.