基于图像处理的高熵合金图像定量分析技术-中国材料进展

文章信息/Info

Title:: Quantitative Analysis Techniques for High-Entropy Alloy Images Based on Image Processing

作者:: 张尊帅; 刘昱含; 郎志远; 周生华; 杨炯; 1. 上海大学材料基因组工程研究院，上海 200444 2. 上海大学钱伟长学院，上海 200444

Author(s):: ZHANG Zunshuai; LIU Yuhan; LANG Zhiyuan; ZHOU Shenghua; YANG Jiong; 1. Materials Genome Institute, Shanghai University, Shanghai 200444, China 2. Qianweichang College, Shanghai University, Shanghai 200444, China

Keywords:: machine learning; scanning electron microscope; image segmentation; image processing; quantitative analysis

摘要:: 扫描电子显微镜（scanning electron microscope，SEM）在材料表征领域具有广泛的应用前景，然而所获得的图像通常难以直接提取定量信息。针对一种共晶高熵合金的扫描电镜图像，提出了一种基于机器学习和图像分割技术的自动化、定量化分析方法，该方法能够有效测量共晶高熵合金板条状区域的面积、长度、宽度、周长以及不同组分的占比。实验结果表明，本研究所提出的方法在高熵合金图像上具有良好的鲁棒性和准确性，为研究高熵合金材料的表面结构提供了重要的技术支持。

Abstract:: Scanning electron microscopes (SEM) hold broad application prospects in the field of material characterization. However, the images obtained are usually difficult to use for direct extraction of quantitative information. This paper proposes an automated and quantitative analysis method based on machine learning and image segmentation techniques for the scanning electron microscopy images of eutectic highentropy alloys. The method effectively measures the area, length, width, circumference and proportion of different components in the lamellar regions of eutectic high-entropy alloys.Experimental results demonstrate that the proposed method in this study exhibits good robustness and accuracy on a large scale of high-entropy alloy images,providing significant technical support for studying the surface structures of high-entropy alloy materials.