Fast particle characterization using digital holography and neural networks.

Area of Science:

• Optical microscopy
• Computational imaging
• Materials science

Background:

• Accurate characterization of coated particles is crucial in various scientific fields.
• Traditional methods for determining core and shell diameters can be time-consuming and may lack precision.
• Existing techniques often categorize particle sizes rather than providing continuous measurements.

Purpose of the Study:

• To develop a rapid method for determining core and shell diameters of coated, non-absorbing spheres.
• To leverage neural networks and digital holographic microscopy for enhanced particle analysis.
• To provide continuous size measurements and differentiate between core and shell dimensions.

Main Methods:

• Integration of a neural network algorithm with digital holographic microscopy (DHM).
• Direct parameter extraction without the need for complex image reconstruction.
• Application to simulated coated spheres with sizes ranging from 7 to 20 μm.

Main Results:

• Achieved high accuracy in determining core diameter: (4.4±0.2)%.
• Demonstrated superior accuracy in determining shell diameter: (0.74±0.01)%.
• Enabled continuous measurement of particle parameters, overcoming limitations of discrete categorization.

Conclusions:

• The proposed neural network-based DHM approach offers a rapid and accurate method for characterizing coated spheres.
• This technique successfully determines both core and shell diameters with high precision.
• The ability to obtain continuous size values represents a significant advancement over previous methods.