<![CDATA[This research aims to develop an automated approach for nanoparticle segmentation within carbon composites using a U-Net model. Nanoparticles in carbon composites are critical for enhancing the mechanical and electrical properties of these materials, but manual detection and segmentation are challenging due to their minute size and dispersed distribution. In this study, a U-Net model with an encoder-decoder architecture was employed to segment scanning electron microscope (SEM) images of palladium-carbon (Pd/C) nanoparticles. The dataset comprised 750 SEM images, exhibiting diverse nanoparticle shapes and sizes. Preprocessing steps included image cropping to eliminate irrelevant regions and the application of Otsu Thresholding to generate ground truth segmentation masks. Model performance was assessed using metrics such as Intersection over Union (IoU), accuracy, and loss. The U-Net model demonstrated high segmentation accuracy, achieving rates between 92% and 95% after 20 training epochs. Additionally, the model was deployed via a Flask web application for real-time prediction. This work significantly advances the efficiency and accuracy of nanoparticle segmentation, offering promising applications in material science and industrial research.]]>
Copyrights © 2024
