Accelerating Microstructure Recognition of Nickel-Based Superalloy Data by UNet++

In recent years, the field of alloys has been developed rapidly under the high-throughput experiments. At the same time, the research and application of superalloy microstructure have become a very important part in the field of alloys. However, it is difficult to deal with the massive images with inconsistent brightness and contrast by conventional methods, which limit the development of superalloy microstructure research. In this paper, combining the traditional threshold segmentation, we propose a microstructure segmentation method based on UNet++, which circumvents the large amount of labeled training data that would need intensive labor. This integrated approach improves efficiency and accuracy compared with traditional methods, and can be applied to many other fields and data.