Abstract
<title>Abstract</title> <p>X-ray diffraction (XRD) analysis that is dependent on laboratory instruments and complete reference databases limits the accuracy of mineral identification in geological fieldwork tasks. Mobile computing provides an on-site decision-making path but deployment of large crystallographic datasets remains constant challenges in storage and query performance for resource-limited devices. In this context, we present MinIDX: an offline Android mineral identification system leveraging a compact crystallographic database and a mobile-friendly search architecture. The XRD reference patterns were obtained from the Crystallography Open Database (COD) using a Python-based preprocessing workflow. The twenty most prominent diffraction peaks of each mineral phase were selected, and the processed data were assembled in a 196 MB SQLite database suitable for use in embedded systems. A configurable indexed fuzzy-matching algorithm accomplishes mineral identification with a d-spacing tolerance of ± 0.02 Å, permitting rapid access to results directly on the mobile device without an Internet connection. Examples were validated on forty-four geological samples across Algeria's major tectonic provinces with a considerable range of lithologies and mineral assemblages, resulting in 100% retrieval of primary mineral phases. Individual diffraction peaks were matched to laboratory XRD measurements within the specified tolerance 92.4% of the time, which also required an average identification time of 2.12 s on mid-range Android hardware. These results show that effective structuring and compressing cryptographic reference data can enable rapid, accurate offline mineral identification on mobile devices, providing a useful reference resource for geological mapping, mineral exploration, education and field work where laboratory facilities or network access to the internet is restricted.</p>