Micrographs are images that encode a visual representation of a material’s microstructure; these are rich in information that can be used for analyzing the processing history of the material as well as predicting its properties and performance. However, the analysis of micrographs has historically been dominated by qualitative observations and most quantitative metrics were developed ad-hoc for a particular material or type of task. This presentation introduces the concept of domain transfer learning coupled with recent advances in computer vision and machine learning as general-purpose tools in micrograph analysis. These tools can be used for a variety of applications such as defect classification, morphology description, and data exploration. Furthermore, these tools can be applied to the development of materials and existing production lines as feedback mechanisms to quantify visual information encoded in micrograph data in an automated workflow.