Machine Learning Inspired Nanowire Classification Method based on Nanowire Array Scanning Electron Microscope Images.

Area of Science:

• Materials Science
• Nanotechnology
• Computer Science

Background:

• Scanning electron microscopy (SEM) is crucial for visualizing nanoscale structures.
• Accurate identification of nanowires in SEM images is essential for material characterization.

Purpose of the Study:

• To develop an innovative classification methodology for identifying nanowires in SEM images.
• To demonstrate the effectiveness of a machine learning (ML)-based approach for nanowire categorization.

Main Methods:

• Utilized advanced image manipulation techniques.
• Employed machine learning-based recognition algorithms for classification.
• Trained models on SEM images of III-V nanowire arrays grown via metal organic chemical vapor deposition.

Main Results:

• Achieved an average F1 score of 0.91, indicating high precision and recall.
• Demonstrated proficiency in isolating and distinguishing individual nanowires within arrays.
• Successfully detected parasitic crystals alongside nanowires.

Conclusions:

• The ML-based method offers high accuracy and performance for nanowire identification.
• The technique is viable for both academic research and practical commercial applications.
• This approach enhances the analysis of SEM images in nanotechnology.