As an important auxiliary means to improve polarization detection technology, infrared polarization simulation technology can provide a theoretical basis and reference for the design of infrared polarization detectors. For imaging simulation, the most accurate method is to use the ray tracing method and the idea of global illumination to simulate all the energy interaction between all rays and the surface. However, for polarization imaging simulation, based on the original massive calculation, there is more polarization state transmission process, which is disastrous for most projects. To solve the problems of complicated calculation of traditional polarization bidirectional reflection distribution function and poor real-time rendering, based on the microplane theory, a faster polarization bidirectional reflection distribution function model is proposed, and the imaging simulation of the whole link is completed. In this research, a semi-empirical model is used to simulate the shadowing and shading effects of radiant energy on rough surfaces using mathematical modeling. It avoids building a 3D model at the micro-facet scale, greatly reducing the workload of creating the sea surface and subsequent rendering. The sea surface is generated according to the P-M wave spectrum, and the Cox-Munk model is used to calculate the slope variance sigma(2) of the sea surface, which is abstracted as the material properties of the sea surface, which reduces the complex calculation process and still conforms to the objective physical laws. A three-dimensional data storage structure suitable for polarization simulation is designed. For infrared polarization simulation, the illumination and color data of the model are not required, but for different material modules, material properties such as complex refractive index and roughness need to be added. Therefore, the vertices and surfels in the original ASE file are re-divided into modules according to the ship parts and materials where the surfels are located, and the vertices and surfels in each module are regrouped and numbered. The effective radiation received by the detector is discussed and the radiation control equation is established. For the specific scene of sea surface detection, the effective radiation that the detector is capable of receiving is analyzed, and a relatively complete radiation control equation is established. The created simulation model is more realistic. For infrared detection, the spontaneous radiation of the target and the sea surface is also a non-negligible part of the energy received by the detector. According to Kirchhoff's law, the spontaneous radiation polarization model of the sea surface and the ship target is established. It is assumed that the average orientation of micro-surface elements is represented by the intermediate vector between the macro-surface element normal and the detection direction, the traditional PG polarization bidirectional reflection distribution function model is improved, and a polarization bidirectional reflection distribution function model that is more suitable for computer real-time rendering is proposed, which balances the Authenticity and real-time requirements of simulation. The directional hemispherical polarization reflectivity and emissivity models of the sea surface and ships are established. Finally, the detector is modeled, and the radiance and polarization state of each surface element is calculated at the same time. The focal length, aperture, responsivity, and other parameters of the detector are used to establish a preliminary detector model, carry out reasonable grayscale mapping, generate S-0, S-1, and S-2, images, and calculate the polarization degree map to complete the simulation work of the whole link. The images of ships on the real water surface under similar conditions were collected and compared with the simulated images, and the gray-scale distributions of the polarization images were similar. The time-consuming results of imaging simulation of sea surface targets respectively show that compared with the traditional model, the simulation model improves the speed of imaging simulation under the premise of ensuring the correctness of the model.T he simulation results provide theoretical support and data basis for target recognition of ships on the sea surface, wind speed inversion of sea surface remote sensing images, and feasibility demonstration before actual detection.
