Algorithms for calculating polarization direction based on spatial modulation of vector optical field

Abstract

Polarization is an important property of electromagnetic waves, and measuring their polarization properties fast and precisely is a very important issue in many applications, such as skylight polarization navigation, optical activity measurement, imaging polarimetry, spectroscopic ellipsometry, fluorescence polarization immunoassay and so on. The polarization measurement method based on vector optical field modulation and image processing is a new type of spatial modulation polarization detection technology. The key step of this technique moving to practical application is determined by the effective polarization measuring algorithms with high speed and accuracy. In order to find out the fast and high precision polarization direction calculation methods, the principle of polarization direction measurement based on vector optical field and spatial modulation is introduced briefly, and the basic characteristics of the spatially modulated intensity distribution images are analyzed. According to the properties of spatially modulated image, we have designed and implemented four different polarization direction calculation methods, which are the Radon transform, intensity modulation curve detection, radial integration and image correlation detection, and their working principles and physical thoughts are introduced elaborately. To compare the detailed performance of these four algorithms, an experimental setup has been constructed to collect the images and perform the algorithm verification, and the stability, speed and accuracy of the four algorithms are compared respectively. The research results indicate that all the four methods can achieve stable and reliable polarization direction detection. The three methods of intensity modulation curve detection, radial integration and image correlation detection, can obtain the measurement accuracy of polarization direction better than 0.01 degrees. The intensity modulation curve detection and radial integration own the fast calculation speed relatively, and their overall performances are the best, and they are the most promising methods for realizing real-time and high-precision polarization measurement.