Resolution Enhancement for Drill-Core Hyperspectral Mineral Mapping

Drill-core samples are a key component in mineral exploration campaigns, and their rapid objective analysis is becoming increasingly important. Hyperspectral imaging of drill-cores non-destructive technique that allows for non-invasive fast mapping phases alteration patterns. The use adapted machine learning techniques such as supervised algorithms robust accurate drill-core hyperspectral data. One the remaining challenge spatial sampling sensors operational conditions, which does not allow us to render textural diversity required map minerals with low abundances fine structures veins faults. In this work, we propose methodology implement resolution enhancement technique, coupled non-negative matrix factorization, using hyperspectral, RGB images high-resolution mineralogical data produce maps at higher resolutions improve minerals. results demonstrate enhanced only provide better details patterns but also were previously hidden due its sampling.