Fourier ptychography for high-resolution imaging has been a revolutionizing technical, since it can provide abundant information about target scene by changing illumination or pupil scanning. However, many objects are covered by dynamic scattering media, such as biological tissues and mist, that disrupts the light paths and forms the scattering wall, let alone high-resolution imaging. It is worth noting that the scatting effect caused by the scattering media will reduce the correlation of scattered light field, which makes the information aliasing difficult to extract. The situation becomes worse if the image scene is in color. Typically, the wavefront shaping, optical transmission matrix, and speckle correlation technique can successfully recover hidden targets form the scattered light field. Notably, the physical model of conventional method is limited by the difficultly in extracting target information from the strong scattering environment, especially in broadband light illumination imaging. Thus, it is limited to achieve super-resolution color imaging through scattering media by utilizing the current techniques.

In this work, we present a computational polarized colorful Fourier ptychography imaging approach for super-resolution perspective in broadband dynamic scattering media. In order to address the challenge of current imaging methods that is limited by the width of the light spectrum, the polarization characteristics of the scattered-light-field are explored. After retrieving a series of sub-polarized images, which bring the information about different frequencies caused by the motion of scattering media and are processed by the common-mode rejection of polarization characteristic, our computational approach utilizes the iterative optimization algorithm to recover the scene. Notably, owning to the difference between the target scattering information and background scattering information of scattered light fields with different polarization rotation angles, we can obtain two images in which the target information and the background information are dominant in the scattered field. Afterwards, a series of images containing target information and background information is used to iterate the Fourier ptychographyprogram to update the target image based on the obtained image sequence until the estimation converges. During the updating procedure, the scattering effect can be removed, and the spatial-resolution is improved.

Compared with traditional scattering imaging model, the proposed method can perform super-resolution color imaging and descattering under various conditions, and solve the problem of color cases. Furthermore, the proposed method is easy to incorporate into a traditional Fourier Ptychography imaging system to obtain high-fidelity images with better quality and effective detail information. Therefore, the proposed method has the potential to help super-resolution imaging to obtain more practical applications.