Pixelated Polarization Imaging Technique And Its Application In Digital Holographic Microscopy

Polarization imaging is intended to measure the polarization information described by Stokes parameters and its derivatives.It has a wide range of applications in image contrast enhancement,remote sensing,biomedical diagnosis,optical rotation measurement,photoelasticity,holography,and interferometry.Pixelated polarization camera has advantages of simple,compact,and integrated structure and high temporal resolution,thus it's gradually becoming the mainstream choice for the measurement of polarization information in these fields.This paper focuses on basic and critical topics in the pixelated polarization imaging technique,including the establishment of pixel models,the experimental calibration of model parameters,the estimation of Stokes parameters,the evaluation of systematic errors and random errors,the pixelated polarization phase shifting technique,aims to provide the theoretical and experimental guidance for the practical uses of pixelated polarization camera.The main contents of this paper are as follows:(1)A relatively perfect pixel model was established to characterize the Stokes parameters measurement of the pixelated polarization camera.(2)Existing interpolation-based and frequency domain filtering-based Stokes parameter estimation methods were summarized,and the convolutional implementations of nearest-neighbor,bilinear,bicubic,and natural bicubic spline interpolation for the Stokes parameters estimation are given.Two general estimation methods,local leastsquares and global smoothing regularization estimation methods,are proposed,which can be used to estimate Stokes parameters under non-uniform modulation.(3)Systematic errors caused by the difference of the instantaneous field of view of the pixels and the non-uniformity of the Stokes parameters responses were theoretical and experimental studied.The equivalent frequency domain filtering implementation of the convolutional interpolation-based estimation method,the local least-squares estimation method,and the global smoothing regularization estimation method under the ideal modulation is derived through Fourier analysis.The abilities of several estimation methods for tackling the instantaneous field of view errors and non-uniformity errors are evaluated and compared.The numerical calibration methods widely used in the existing studies were explained.It was noted that additional instantaneous field of view errors will be introduced when using the superpixel calibration method to tackle the non-uniformity.An experiment method based on the half-wave plate rotation was proposed,which can quantitatively evaluate the estimation errors of the Stokes parameters for the actual pixelated polarization camera.(4)Random errors caused by noise are theoretical and experimental studied.The relationship between systematic errors,random errors and total errors of the Stokes parameters estimation in pixelated polarization imaging technique was clarified for the first time.The random error weights of noise in the Stokes parameters estimation meth-ods were theoretically analyzed and experimentally verified.The effects of the quantum efficiency,gain,noise,primary transmittance,and secondary of the pixel polarization camera on random errors are discussed comprehensively.(5)The pixelated polarization phase shifting technique was explained from the per-spective of polarization imaging,and the advantages and disadvantages between this technique and off-axis spatial phase shifting technique were compared.The condition,which the magnification and numerical aperture of objective,laser wavelength,and the sampling interval of pixelated polarization camera should satisfy to make the microscope system reach diffraction-limited resolution,was given.A digital holographic microscope system based on the pixelated polarization camera was constructed.The numerical focusing function was realized,and the lateral resolution of the system was verified.