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

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, and fluorescence polarization immunoassay . 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 effective polarization measuring algorithms with high speed and accuracy. In order to find out the method of fast and precisely calculating polarization direction, 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 propose and implement four different polarization direction calculation methods, which are the Radon transform, intensity modulation curve detection, radial integration, and image correlation detection, and also introduce their working principles and physical thoughts elaborately. To compare the detailed performances of these four algorithms, an experimental setup is constructed to collect the images and perform the algorithm verification, and the stabilities, speeds and accuracies of the four algorithms are compared. The research results indicate that all the four methods can achieve their stable and reliable polarization direction detections. The three methods, intensity modulation curve detection, radial integration and image correlation detection, can achieve the polarization direction measuring accuracy better than 0.01° . The intensity modulation curve detection and radial integration have relatively fast calculation speed and the best comprehensive performances , and are the most promising methods to realize real-time and high-precision polarization measurement.