A deep learning approach for complex microstructure inference

Abstract

Segmentation and classification of microstructures are required by quality control and materials development. The authors apply deep learning for the segmentation of complex phase steel microstructures, providing a bridge between experimental and computational methods for materials analysis.