Polarization 3D imaging method can directly measure the normal vector of the entire field through a single camera, and has been widely used in many fields. The existing polarization models typically assume that all light reflected from the target surface to the camera is parallel to the camera's optical axis, which is inconsistent with the actual situation and leads to system errors. To address this issue, this paper proposes a polarized 3D imaging pinhole imaging model that can accurately describe the imaging optical path. The model first considers the incident direction of light and analyzes the relationship between the normal vector, reflected light component, and reflectivity; Reuse polarization degree and polarization angle to calculate the reflectivity and the angle between the reflected light component and the imaging plane; Finally, the equation between the normal vector and the reflected light component and reflectance was solved using the inverse method, achieving accurate measurement of the normal vector. The accuracy of this method has been verified through a series of experiments. The results indicate that compared with traditional models, this model has advantages such as wider application range and higher measurement accuracy.