Demosaicking DoFP images using Newton's polynomial interpolation and polarization difference model.

Area of Science:

• Optics and Photonics
• Image Processing
• Sensor Technology

Background:

• Division of focal plane (DoFP) polarization imaging is advancing due to mature nanofabrication.
• Real-time DoFP polarimeters face spatial resolution trade-offs, leading to instantaneous field of view (IFoV) errors.
• Interpolation is crucial for reconstructing missing polarization data in DoFP systems.

Purpose of the Study:

• To present a novel interpolation technique for DoFP polarimeter demosaicking.
• To address spatial resolution limitations and IFoV errors in real-time polarization imaging.
• To enhance the accuracy of polarization information recovery.

Main Methods:

• Developed a Newton's polynomial interpolation technique for DoFP demosaicking.
• Performed interpolation in the polarization difference domain, considering interpolation error.
• Utilized an edge classifier based on polarization difference and a fusion scheme for boundary feature recovery.

Main Results:

• The proposed method effectively reduces nonconformities caused by high-frequency energy.
• Demonstrated superior performance compared to state-of-the-art techniques in both synthetic and real DoFP images.
• Achieved quantitative and visual improvements in visible and long-wave infrared spectrum images.

Conclusions:

• The novel interpolation method significantly enhances the accuracy of DoFP polarimeter demosaicking.
• The technique offers a viable solution for improving real-time polarization imaging quality.
• The approach effectively recovers boundary features and reduces image artifacts.