Lamellar and Bi-Modal Ti-64 Microstructure Images: 40 Optical Micrographs

A collection of 40 optical micrographs of Titanium-64, taken from different regions of a forged billet. The images comprise two distinct morphologies, with 20 images each of bi-modal and lamellar structures. This dataset was compiled by Michael White at the University of Manchester to test machine learning algorithms for microstructure classification.

Use Cases

• Classifying titanium alloy microstructures based on the described bi-modal and lamellar morphologies.
• Training image recognition models to distinguish between different material phases in optical micrographs.
• Benchmarking machine learning algorithms for materials science image analysis tasks.

Strengths

• Contains 40 labeled images, providing a clear basis for binary classification tasks.
• Images are explicitly divided into two equal sets of 20 images for each morphology (bi-modal and lamellar).
• The dataset has a specific, stated purpose for testing machine learning algorithms.

Limitations

• Column-level documentation is absent; image metadata and preprocessing details must be inferred after download.
• Row count is unknown, which may limit suitability assessment for larger-scale model training.
• Last update date is unknown; freshness unverified.

Provenance

Source

University of Manchester

Collection Method

Optical micrographs taken from different regions of a forged Titanium-64 billet.