Semi-automatic image analysis of particle morphology of cellulose nanocrystals

Morphology analysis of cellulose nanocrystals (CNCs) using transmission electron microscopy (TEM) and atomic force microscopy (AFM) images is an important step in the design and optimization of the processes employed in the manufacture and utilization of CNCs. Current protocols used in the analyses of CNC particle morphology for such microscopy images are largely manual and time-consuming, and often produce inconsistent results between different researchers. This paper describes a new semi-automated image analysis framework that can reliably and quickly detect and classify CNCs from TEM and AFM images and measure their dimensions. The proposed image analysis framework is named CNC-Standardized Morphology Analysis for Research and Technology (SMART). The viability of this framework is demonstrated in this paper using exemplar images obtained for a National Research Council Canada certified reference material, CNCD-1. The results obtained from the SMART approach presented in this work are compared critically against the results obtained from the conventional manual approaches. These comparisons revealed a good agreement between the manual and SMART approaches. Notably, the SMART approach showed significant potential for consistent CNC identification and dimensional measurements at a much higher throughput (e.g., number of CNCs measured and number of images analyzed) compared to the conventional manual approaches.