Particle displacement refinement based on hybrid cross-correlation optical flow method with gradient constancy...

Area of Science:

• Fluid dynamics
• Optical measurement techniques
• Image processing

Background:

• Particle Image Velocimetry (PIV) is sensitive to image brightness changes caused by laser fluctuations.
• Traditional optical flow methods (OFM) fail with intensity variations and large displacements in PIV.
• Existing variational OFM lacks robustness and accuracy for PIV applications with varying illumination.

Purpose of the Study:

• To develop a robust optical flow method for PIV that handles large displacements and intensity variations.
• To improve the accuracy and spatial resolution of displacement measurements in PIV.
• To address the limitations of traditional OFM in dynamic fluid flow analysis.

Main Methods:

• Developed a hybrid cross-correlation optical flow method (CC-OFM).
• Integrated brightness and gradient constancy assumptions in the data term to compensate for intensity changes.
• Evaluated CC-OFM using synthetic and experimental particle image data.

Main Results:

• CC-OFM demonstrated high accuracy and robustness for PIV images with large displacements and intensity variations.
• The method effectively compensated for illumination changes between image pairs.
• CC-OFM achieved superior performance compared to other displacement measurement techniques.

Conclusions:

• The proposed CC-OFM significantly enhances the reliability of PIV measurements under challenging conditions.
• High spatial resolution of CC-OFM enables detailed capture of flow dynamics.
• CC-OFM offers a robust solution for analyzing fluid flows with significant illumination variations.