Rodlike nanoparticle parameter measurement method based on improved Mask R-CNN segmentation

Parameter measurement of nanoparticle, which aims at evaluation of the quality of nanomaterials, is essential to nanotechnology and many applications. According to the nanoparticle images captured by transmission electron microscopy, this paper presents an automated procedure that can expedite the parameter measurement of the rodlike nanoparticles. Nanoparticle segmentation is the most important step in nanoparticle parameter measurement. The challenge of this task involves segmenting the adhesive nanoparticles and nanoparticles with weak contours. To accurately measure nanoparticle size and evaluate nanomaterial quality, firstly, according to the characteristics of agglomeration and adhesion of nanoparticle images, the Mask R-CNN network was selected to segment the nanoparticle images, and the network was optimized to improve the segmentation accuracy. Secondly, according to the particle segmentation result, the minimum circumscribed rectangle of the rodlike nanoparticle boundary is obtained. Finally, the size and shape parameters of the particles are measured based on the minimum circumscribed rectangle. The experimental results confirm the effectiveness of the proposed method for measuring the rodlike nanoparticle parameters.

Automated summary

This paper presents an automated procedure for expedited parameter measurement of rodlike nanoparticles, including key steps such as nanoparticle segmentation. By using the Mask R-CNN network to segment nanoparticle images and optimizing the network to improve accuracy, the size and shape parameters of the nanoparticles were successfully measured.