Clustering Algorithms for Geological Image Categorization

The contract seeks the design and implementation of clustering algorithms to categorize segmented petrographic images based on mineral composition, texture, and structural patterns, with the goal of enabling accurate rock type classification. This effort is critical for advancing automated geological analysis by transforming visual data into quantifiable classifications that support downstream geological modeling and resource assessment. The work must produce robust, scalable algorithms capable of handling high-resolution imaging datasets typical of petrographic studies, ensuring reliable grouping even under variable lighting, staining, and sample preparation conditions. The contract is a subcontract issued by the Government of Canada through the Department of Natural Resources, with a NAICS code of 541721 indicating it falls under scientific research and development services. The solicitation was posted on July 23, 2026, and responses are due by August 7, 2026. Performance is to be carried out within Canada, and while no specific location or point of contact is provided, the project is tied to national geological research priorities. The work does not involve a set-aside program, and applicants are expected to deliver algorithmic solutions that integrate seamlessly with existing geological imaging platforms and workflows used by federal agencies.