A Comparison of Different Counting Methods for a Holographic Particle Counter: Designs, Validations and Results

Area of Science:

• 3D Imaging
• Optical Metrology
• Particle Analysis

Background:

• Digital holographic imaging (DIH) is widely used for 3D particle analysis, often requiring complex wavefront reconstruction.
• Observing particle densities, particularly for spherical particles like those in High Performance Computing (HPC) applications, presents a simplified imaging challenge.
• Spherical particles generate circular fringe patterns on the hologram plane, detectable via 2D image processing.

Purpose of the Study:

• To investigate a simplified approach for determining particle number concentrations using fringe pattern analysis in DIH.
• To identify and extract key features from fringe patterns for effective recognition.
• To develop and compare pattern recognition techniques for detecting and counting fringe patterns.

Main Methods:

• Analysis of fringe pattern characteristics at the hologram plane.
• Feature extraction from fringe patterns to guide pattern recognition.
• Development and customization of three distinct fringe pattern detection and counting algorithms.

Main Results:

• Identification of relevant fringe pattern features for pattern recognition.
• Successful implementation of three customized techniques for fringe pattern detection.
• Comparative analysis of the developed techniques based on detection performance and computational efficiency.

Conclusions:

• Particle number concentration can be accurately determined by counting fringe patterns, simplifying DIH applications.
• The extracted fringe pattern features are crucial for selecting and parameterizing appropriate recognition algorithms.
• The presented techniques offer viable solutions for efficient and accurate fringe pattern detection in particle density analysis.