Accessible deep learning for automated segmentation of supported nanoparticles in electron microscopy

Fast and accurate quantification of the size and morphology of nanoparticles in electron microscopy images is essential for advancing heterogeneous catalysis and energy-conversion research, yet manual segmentation remains the main approach, which is time-consuming, subjective, and challenging to scale. Herein, we present an accessible and efficient deep learning model integrated into a fully functional analysis application for rapid segmentation and statistical quantification. Taken together, this work demonstrates that high-quality nanoparticle segmentation in electron microscopy images is feasible even with minimal annotated data and modest computational resources.