The aim of this research is to introduce novel and efficient interpolation algorithms to interpret polarization information for division of focal plane polarization image sensors. Division of focal plane (DoFP) polarization image sensors capture polarization properties of light at every imaging frame. However, these imaging sensors capture only partial polarization information, resulting in reduced spatial resolution output and a varying instantaneous field of overview (IFoV).

To regain missing polarization information for DoFP polarization image sensors, different interpolation techniques are used. The main purpose of interpolation is to present proficient insertion calculations to translate polarization data. The deficiencies should be considered in mind, and the requirement is...[ Read more ]

The aim of this research is to introduce novel and efficient interpolation algorithms to interpret polarization information for division of focal plane polarization image sensors. Division of focal plane (DoFP) polarization image sensors capture polarization properties of light at every imaging frame. However, these imaging sensors capture only partial polarization information, resulting in reduced spatial resolution output and a varying instantaneous field of overview (IFoV).

To regain missing polarization information for DoFP polarization image sensors, different interpolation techniques are used. The main purpose of interpolation is to present proficient insertion calculations to translate polarization data. The deficiencies should be considered in mind, and the requirement is to take a fully favorable advantage of the constant imaging capacities of DoFP image sensors.

We propose algorithms to improve the efficiency of missing polarization information with less error than the state-of-the-art algorithms used for DoFP polarization image sensors. The algorithms judge edges, smoothness and polarization residues to reduce the cross talk of the pixels. We have applied our techniques to different micropolarizer patterns to check which architecture best recovers the missing polarization information, and we have tested them on indoor images, outdoor images and human liver carcinoma tissues for both visual evaluation and mean square error. Our results show that our methods provide high accuracy, low error, low complexity, balance of the state-of-the-art interpolation algorithms, discontinuity reconstruction along edges and increased gain at higher spatial frequencies. This enables processing on a single chip for real polarization image sensors.