Digital Fingerprinting of Microstructures

TL;DR

This work introduces a unified microstructural fingerprinting framework that converts micrographs into compact, information-rich vectors by extracting base features (patch-based or CNN-derived), clustering these features into $K$ centers, and computing order-$l$ statistics $H_0$, $H_1$, and $H_2$ (with multi-scale and CNN-augmented variants). Fingerprints are then used for supervised, semi-supervised, and unsupervised learning, with PCA-based reductions enabling scalable deployment. Key findings show transfer-learning CNN features often outperform traditional keypoint descriptors, Poisson learning offers advantages at very low label rates, and spectral clustering generally improves unsupervised results, while multi-scale and VLAD-type normalizations bolster performance. The framework is validated on Ti alloy and ultrahigh-carbon steel micrographs, suggesting practical utility for rapid high-throughput screening and enabling robust process-structure-property exploration through compact, transferable representations.

Abstract

Finding efficient means of fingerprinting microstructural information is a critical step towards harnessing data-centric machine learning approaches. A statistical framework is systematically developed for compressed characterisation of a population of images, which includes some classical computer vision methods as special cases. The focus is on materials microstructure. The ultimate purpose is to rapidly fingerprint sample images in the context of various high-throughput design/make/test scenarios. This includes, but is not limited to, quantification of the disparity between microstructures for quality control, classifying microstructures, predicting materials properties from image data and identifying potential processing routes to engineer new materials with specific properties. Here, we consider microstructure classification and utilise the resulting features over a range of related machine learning tasks, namely supervised, semi-supervised, and unsupervised learning. The approach is applied to two distinct datasets to illustrate various aspects and some recommendations are made based on the findings. In particular, methods that leverage transfer learning with convolutional neural networks (CNNs), pretrained on the ImageNet dataset, are generally shown to outperform other methods. Additionally, dimensionality reduction of these CNN-based fingerprints is shown to have negligible impact on classification accuracy for the supervised learning approaches considered. In situations where there is a large dataset with only a handful of images labelled, graph-based label propagation to unlabelled data is shown to be favourable over discarding unlabelled data and performing supervised learning. In particular, label propagation by Poisson learning is shown to be highly effective at low label rates.

Figures (6)

• Figure 1: Visual representation of feature population construction and clustering from SIFT keypoint features, where each feature is represented as a ring, centred at a keypoint, with size and orientation corresponding to the scale at which the feature was detected and the dominant orientation of the feature, respectively.
• Figure 2: Schematic of $H_0$ fingerprint construction from SIFT features.
• Figure 3: Example micrographs from LFTi64 dataset.
• Figure 4: Example fingerprints from LFTi64 dataset, constructed from SURF features clustered with 10 cluster centres. These are ordered such that the mean of bi-modal microstructural fingerprints is in ascending order.
• Figure 5: Example microstructures from UHCS600 dataset.