Automated analysis of heterogeneous carbon nanostructures by high-resolution electron microscopy and on-line image processing

High-resolution electron microscopy is an efficient tool for characterizing heterogeneous nanostructures; however, currently the analysis is a laborious and time-consuming manual process. In order to be able to accurately and robustly quantify heterostructures, one must obtain a statistically high number of micrographs showing images of the appropriate sub-structures. The second step of analysis is usually the application of digital image processing techniques in order to extract meaningful structural descriptors from the acquired images. In this paper it will be shown that by applying on-line image processing and basic machine vision algorithms, it is possible to fully automate the image acquisition step; therefore, the number of acquired images in a given time can be increased drastically without the need for additional human labor. The proposed automation technique works by computing fields of structural descriptors in situ and thus outputs sets of the desired structural descriptors in real-time. The merits of the method are demonstrated by using combustion-generated black carbon samples.

► The HRTEM analysis of heterogeneous nanostructures is a tedious manual process.
► Automatic HRTEM image acquisition and analysis can improve data quantity and quality.
► We propose a method based on on-line image analysis for the automation of HRTEM image acquisition.
► The proposed method is demonstrated using HRTEM images of soot particles.