An Extended Polarization Sensor Model Considering the Influence of Obliquely Incident Direct Sunlight

The polarization sensor model plays a key role in bioinspired polarization navigation. Existing models consider the linearly polarized light and vertically incident unpolarized light. The performance of the polarization sensor, however, will be degraded due to the effect of obliquely incident direct sunlight. To account for the influence of obliquely incident direct sunlight and improve the conventional model, we established an extended polarization sensor model. In the proposed model, the transmitted intensity of direct sunlight obliquely passing through the polarizer is analyzed, and the influence of obliquely incident direct sunlight on the polarization sensor is characterized. Then, the extended polarization model is established considering the linearly polarized light and vertically incident unpolarized light, as well as the obliquely incident direct sunlight. Based on the proposed model, the secant Levenberg-Marquardt (LM) least-squares algorithm is used to estimate the polarization parameters of interest, such as Degree of Polarization (DoP) and Angle of Polarization (AoP). Finally, the effectiveness of the proposed polarization sensor model is verified via simulation, indoor, and outdoor experiments. The results show that the root mean square of AoP error can be narrowed to 0.67 degrees even under the obliquely incident direct sunlight environment.