Magnesium Technology 2024
Magnesium Technology 2024
复制标题
镁技术 2024
DOI：
10.1007/978-3-031-50240-8_10
复制
发表时间：
2024
期刊：
影响因子：
--
通讯作者：
Yi H
中科院分区：
文献类型：
--
作者：
Yi H
电子背散射衍射法在金属领域被广泛采用。然而，尽管数据源丰富，但往往缺乏涵盖所有特征的充分分析。特别是随着新兴的原位技术，数据处理非常耗时，因此必须访问每一位数据。在这项工作中，开发了一个工具包，旨在自动有效地处理 EBSD 数据。工具包的两部分是用Matlab和Mtex开发的。一种用于关联两张图，实现简单，几分钟内即可生成结果，表明两张图之间的颗粒相关性。另一个将一系列现场数据集关联起来，使每个单独的颗粒都变得可追踪。在工具包的帮助下，可以通过原位过程创建包含像素、数字信息和颗粒属性的大型数据集。因此，使用新颖的数据科学方法，特别是机器学习和深度学习来研究微观特征和颗粒行为。
Electron backscatter diffraction method is widely adopted in metal fields. However, despite the abundant data sources, sufficient analysis covering all features is often absent. Especially with the emerging in-situ techniques, data processing is time-consuming, where access to every bit of data is imperative. In this work, a toolkit is developed with the aim of processing EBSD data automatically and efficiently. Two parts of toolkits are developed with Matlab and Mtex. One is used to correlate two maps, with simple implementation, results will generate within few minutes, indicating the grains correlation between two maps. The other correlates a series of in-situ datasets, making each individual grain become trackable. With the assistance of the toolkits, a large dataset containing pixels, digital information, and grains properties through an in-situ process can be created. Thus, the microfeatures and grain behaviors are studied using novel data science methods, especially machine learning and deep learning.