Wójcicki Tomasz, Nowicki Michał
Institute for Sustainable Technologies, National Research Institute, Pulaski Str. 6/10, 26-600 Radom, Poland.
Industrial Research Institute for Automation and Measurements PIAP, Jerozolimskie 202, 02-486 Warsaw, Poland.
Materials (Basel). 2016 Apr 1;9(4):265. doi: 10.3390/ma9040265.
The article presents a selected area of research and development concerning the methods of material analysis based on the automatic image recognition of the investigated metallographic sections. The objectives of the analyses of the materials for gas nitriding technology are described. The methods of the preparation of nitrided layers, the steps of the process and the construction and operation of devices for gas nitriding are given. We discuss the possibility of using the methods of digital images processing in the analysis of the materials, as well as their essential task groups: improving the quality of the images, segmentation, morphological transformations and image recognition. The developed analysis model of the nitrided layers formation, covering image processing and analysis techniques, as well as selected methods of artificial intelligence are presented. The model is divided into stages, which are formalized in order to better reproduce their actions. The validation of the presented method is performed. The advantages and limitations of the developed solution, as well as the possibilities of its practical use, are listed.
本文介绍了基于对所研究金相截面的自动图像识别进行材料分析的选定研发领域。描述了气体氮化技术材料分析的目标。给出了渗氮层的制备方法、工艺步骤以及气体氮化设备的构造和操作。我们讨论了在材料分析中使用数字图像处理方法的可能性,以及它们的基本任务组:提高图像质量、分割、形态变换和图像识别。介绍了所开发的渗氮层形成分析模型,涵盖图像处理和分析技术以及选定的人工智能方法。该模型分为多个阶段,这些阶段被形式化以便更好地重现其动作。对所提出的方法进行了验证。列出了所开发解决方案的优点和局限性以及其实际应用的可能性。