Clustering of polarization-encoded images

Area of Science:

• Optics and Photonics
• Image Processing
• Computational Imaging

Background:

• Polarization-encoded imaging captures rich polarization information per pixel.
• Analyzing multidimensional polarization data presents significant challenges.
• Understanding spatial coherence is key to advanced image analysis.

Purpose of the Study:

• To develop and validate novel algorithms for clustering multidimensional polarization-encoded images.
• To investigate the role of spatial coherence in polarization information.
• To improve the analysis of data from fully polarimetric imaging systems.

Main Methods:

• Proposed two analysis methods: polarization contrast enhancement and a Markovian model-based image processing algorithm.
• Considered the spatial coherence of polarization information.
• Applied and validated algorithms using Mueller images from a polarimetric imaging system.

Main Results:

• Demonstrated the effectiveness of the proposed algorithms in clustering multidimensional polarization images.
• Showcased improved analysis capabilities through polarization contrast enhancement.
• Validated the Markovian model approach for sophisticated image processing.

Conclusions:

• The developed algorithms provide effective tools for analyzing complex polarization-encoded images.
• Spatial coherence analysis is a valuable approach for enhancing polarimetric image interpretation.
• The methods are applicable to data from various polarimetric imaging systems.