Recently several polarimetric techniques have been suggested, designed deliberately for automatic whole-field birefringence imaging in photoelastic models with essentially three-dimensional stresses. In general, these techniques are feasible for mapping three optical parameters that determine birefringence in a given case. However, the difficulty in attaining a high level of data accuracy over the whole image persists. There remains a problem of precise imaging in regions where the mutual interference of three given parameters inevitably causes accuracy deterioration. We show how to correct such imperfections in an imaging polarizer–sample–analyzer (PSA) Fourier polarimetry technique, as suggested earlier [Appl. Opt. 41, 644 (2001)]. The given technique (a method developed so that it maps the phase, the azimuth, and the ellipticity angles of an elliptic retarder) particularly fails to provide precise imaging in regions where the phase is either close to null or approaches π-multiple values and in intervals where the ellipticity angle falls into the proximity of ±π/4 values. These drawbacks can be successfully overcome by incorporation of a compensator into a PSA polarimeter arrangement. Although use of a compensator in the polarimeter makes the original technique more complicated, we demonstrate that the compensator allows two important issues to be resolved. First, it provides precise imaging for each of three optical parameters through the whole accessible intervals of the parameters regardless of the absolute value of the parameter. In addition, it gives a sign of phase that remains undefined in the PSA techniques. Theoretical considerations are presented and are followed by experimental data that illustrate the improved accuracy capabilities of the compensator-enhanced technique.
© 2001 Optical Society of AmericaFull Article | PDF Article