Fig. 1 Image processing using the Hessian matrix76Persistent homology is a powerful tool for quantifying various structures, but it is equally crucial to maintain its interpretability for material design. In this study, we extracted interpretable geometric features from the persistent diagrams (PDs) of scanning transmission electron microscopy (STEM) images of self-assembled Pt-CeO2 nanostructures synthesized under different annealing conditions. Analysis of the PD quadrants provided five interpretable features: average width and total length of striped CeO2 phases, the number of CeO2 phases from zeroth PDs, and the numbers of ring- and arc-like structures from first PDs. Principal component analysis (PCA) and its component mapping onto PDs clarified that the number of small arc-like structures is especially important for describing Pt-CeO2 nano- structural changes. This descriptor enabled us to quantify the degree of disorder, namely the density of bends, in nanostructures formed under different conditions. By using this descriptor along with the width of the CeO2 phase, we could classify 12 Pt-CeO2 nanostructures in an interpretable way [1]. Rubber’s unique elasticity and viscoelasticity among materials make it suitable for various applications, including tires and medical devices. These excellent properties are attributed to the crosslinking structure that connects the molecular chains [1]. However, significant noise makes it difficult to accurately understand the internal crosslinking structure of rubber, even with electron microscopy. This research focuses on improving the visualization of crosslinked network structures in rubber using image processing techniques. The study utilizes the Hessian matrix, a method commonly used in bio-medical imaging [2], to enhance the visibility of the network structure, even in noisy and hazy images. This allows for extracting parameters related to the network and its density, which correlates strongly with experimental results. This method successfully compared the internal structure of crosslinked rubbers of different formulations.Poster Award NomineePoster Award NomineeImage Processing Techniques for Unveiling the Internal Structure of Crosslinked RubberInterpretable Structural Evaluation of Metal-Oxide Nanostructures in STEM Images via Persistent Homology P4-09Ryuto Eguchi1,2, Yu Wen1,2, Hideki Abe1,3, and Ayako Hashimoto1,2 1 Research Center for Energy and Environmental Materials, National Institute for Materials Science (NIMS) 2 Graduate School of Science and Technology, University of Tsukuba 3 Graduate School of Science and Engineering, Saitama University [1] R. Eguchi et. al., Nanomaterials, 14, 1413 (2024). P4-10Masato Suzuki1,2 and Yasuhiko Igarashi1,31 Institute of Systems and Information Engineering, University of Tsukuba2 The Yokohama Rubber Co., Ltd.3 Center for Basic Research on Materials, National Institute for Materials Science (NIMS)[1] D. Y. Kim et al., Polymers, vol. 12, no. 9, 2020 (2020).[2] B. P. Marsh et al., Sci Rep, vol. 8, no. 1, 978 (2018).
元のページ ../index.html#76