Imaging through scattering media, such as clouds, biological tissues, and seawater, has broad application prospects in transportation, medical diagnosis, and information technology. Researchers have proposed various techniques to obtain images from scattered light passing through the scattering media, among which speckle correlation imaging has developed rapidly. Speckle correlation imaging requires non-coherent light sources due to the requirement of memory effect. This requirement makes the imaging device complex, and the light source utilization low. Additionally, this method is limited in its application under the illumination of spatially coherent light sources. This paper proposes a new method of speckle correlation imaging based on the illumination of spatially coherent light, which is achieved by multiplexing different polarization direction speckle patterns, called polarization multiplexing scattering imaging. To achieve decoherence of the light source, previous approaches have used a rotating scattering medium to generate time-varying speckle patterns that are integrated over the shutter time of the camera to eliminate coherent noise; or multiplexed wavelength-dependent speckle multiplexing to achieve this. This paper uses spatially incoherent light sources to obtains different polarization direction speckle patterns by rotating polarizers placed in the illumination path. These patterns are superimposed and averaged, and phase recovery algorithms are used to reconstruct the object image. This experiment uses Ping-Pang (PP) algorithm with fusion error reduction and hybrid input-output algorithm to quickly and high-quality reconstruct targets. By comparing the reconstruction results of different numbers of reused speckle patterns, it is shown that using more speckle patterns can achieve better image quality. Compared with conventional speckle correlation imaging technology, the proposed method reduces the requirements for light sources, improves the utilization rate of light sources, and makes the device simpler and more compact. Experimental results show that this method is feasible and has strong environmental adaptability, which can expand the application range of speckle correlation imaging methods.
