On the occasion of World Metrology Day, whose theme this year was “Metrology in the digital age”, LNE (French National Laboratory of Metrology and Testing) presented the NanoMetrologIA platform, a tool for measuring the dimensional properties of nanoparticles using artificial intelligence.
The adoption of digital technology is revolutionising metrology by improving processes and opening up new opportunities. This is particularly the case for the characterisation of nano-objects, and more particularly nanoparticles.
The characterisation of nanoparticles is of prime importance to the industry. It is in this context that LNE has developed the NanometrologIA platform.
In practice, the measurement of the size of nanoparticles is based on multiple images obtained by electron microscopy, which make it possible to produce a statistical distribution of nanoparticle diameters.
Accelerating the measurement of nanoparticles
Conventional methods, which are essentially manual, require several hours of processing and only work on particles whose entire contour is visible in the image. They are not able to discern particles in the form of aggregates or agglomerates where they may be partially masked.
To improve analysis and reduce image processing time, LNE has developed an algorithm based on deep learning from a dedicated database. The technique used automatically identifies all isolated or aggregated particles and reconstructs the contour of each particle. Thanks to the NanoMetrologIA platform, the measurement of the size distribution of nanoparticles now only takes a few seconds.
The platform is currently used mainly for the identification of titanium …
