When optical imaging in turbid media, the target signal is drowned in the background noise due to the influence of media absorption and scattering, resulting in a serious degradation of image clarity. As a representative technique of polarization imaging, polarization differential can significantly improve the detection capability of targets in turbid media by using the polarization information of light, which fully exploits the difference in polarization characteristics between medium light and scene light to achieve polarization common-mode suppression of backscattered noise. Due to its simple system structure, good portability, wide range of applicable fields, highlighting the edge and other advantages have attracted much attention. This technology can not only be used as a physical optical imaging method for high contrast imaging of targets in underwater turbid environments, biological tissues, and atmospheric haze scattering environments, but also as an effective means to suppress stray light from optical sensor devices, and a large number of excellent research results have been accumulated since its introduction. From the basic principle of polarization differential imaging technology, according to the different imaging application scenarios, a series of recent research advances in polarization differential imaging is introduced from two aspects. And an outlook on polarization differential imaging is provided, which provides guidelines for subsequent research in this field.

Polarization imaging is a novel photo detection method. Compared with traditional intensity imaging, polarization imaging can further obtain the polarization characteristics of the target, therefore improving the contrast and increasing the recognition probability. Polarization imaging has important application value in biomedical, industrial inspection, earth remote sensing, modern military, marine and aviation fields. The imaging methods of polarization imaging systems are classified and compared, and the realization methods of polarization imaging are summarized. On this basis, according to the characteristics of different polarization imaging systems and different application backgrounds, the future development direction of polarization imaging systems is prospected, including new polarization imaging systems based on metasurfaces and new polarization imaging spectroscopy systems.

Polarization imaging technology deeply excavates the light intensity， spectrum and polarization information of light wave， and synthetically considers the formation mechanism of haze image， and combines atmospheric scattering imaging models to estimate target multi-dimensional physical information to effectively restore real scene images. At present， polarization dehazing technology has been extensively used and leading research results in many fields such as intelligent transportation and security monitoring. The basic principle， physical model， algorithm technology and imaging effect of polarization dehazing technology are systematically introduced， and polarization dehazing technology is analyzed and prospected according to the advantages and disadvantages of the existing technologies.

Depth information can be acquired via different kinds of methods, like stereo vision, time-of-flight imaging and structured light. But all those three-dimensional (3D) imaging techniques have limitations in practical applications. Polarization is a significant property for light wave which can provide additional detailed information. Applying the polarization characteristics to 3D imaging, namely polarization-based 3D imaging, can make up for the shortcomings of existing 3D imaging methods to a certain extent. The research progress in polarization-based 3D imaging is overviewed based on the various methods of depth measurements and the implementation of polarization character, including polarized stereo vision, polarized time-of-flight imaging and polarized structured light. The advantages and shortcomings of the polarization-based approach are analyzed and summarized, which provides guidance for future research directions in related fields.

According to the requirements of polarization detection and recognition for space targets, the polarization spectral reflection characteristics of six typical space target materials such as gold insulating film, silver insulating film, monocrystalline silicon, gallium arsenide, satellite coatings SR107 and S781 are investigated. The testing results of material polarization spectrum are analyzed, and the spatial distribution of polarization spectrum characteristics of space target is disclosed. The results show that the spectral characteristics don′t change with the detection angle, the spectral peak doesn't change with the polarization analyzer angle. The half-peak width, characteristic wavelength, polarization retaining and depolarization properties of the materials don′t change with the incidence angle, but the values of characteristic spectral peak at different incidence angles are different. The work may provide guidance for polarization detection of space targets.

As an efficient optical detection technology, polarization imaging obtains the polarization information of the target scene through polarization modulation, and then realizes the multi-dimensional target representation, recognition and detection. However, in practice, polarization imaging technology is very easy to be affected by noise, which leads to the degradation of image quality. Especially in the calculation of degree of polarization and angle of polarization, the nonlinear calculation will amplify the noise, then obtain the wrong polarization information, which greatly limits the further application and development of polarization imaging technology. Effective denoising technology can improve the quality of polarization images and obtain more accurate polarization information from polarization images, which has important application value and research significance in practice. Based on the research progress and development status of polarization imaging denoising technology, the basic principles, implementation approaches and algorithms of different techniques are mainly introduced, and their performances are evaluated and compared. The work is of great significance for obtaining high quality polarization parameters and the subsequent polarization information processing and application.

Aiming at the problems that the traditional detection method can′t detect and see the target clearly, and the image contour and detail are blurred. The combination of infrared and polarization detection is adopted to solve the problem that the infrared image information and polarization image information cannot be detected and seen in various environments. Aiming at the problems of large amount of data and slow extraction speed in the process of target local feature extraction, an improved deep learning local feature extraction (SIFT) algorithm for polarization images is proposed. Experimental results show that the improved algorithm combines the advantages of polarization imaging and deep learning to achieve rapid feature extraction of targets in simple or complex backgrounds. This algorithm improves the speed and accuracy of local feature extraction of polarized images. The improved algorithm lays a theoretical foundation for target classification, recognition and tracking technology.

Improving the anti scattering medium interference ability of photoelectric imaging system is always a very challenging and important subject in optics. In scattering media, the image quality is often seriously degraded due to the scattering and absorption effects of small particles and the resulting uneven distribution of intensity or polarization characteristics. Based on the polarization characteristics of atmospheric light, polarization detection can estimate the intensity level and transmittance of atmospheric scattering, and separate the target signal from the interference signal to achieve high-quality imaging. The principle of depolarization based on polarization is first described, and then the latest progress of various polarization imaging systems based on this principle is introduced, such as polarization imaging systems based on polarization difference, Stokes vector and Muller matrix. Finally, the future development of polarization imaging in scattering media is prospected.

Aiming at the Stocks measurement model of continuous rotation polarization imaging system, the factors that affect the model are analyzed, such as initial exposure angle error, uneven rotation speed of polarizer, optical axis deflection, and puts forward the imaging design constraints of a quasi-real-time polarization detection system based on continuous rotation. Combined with the actual situation, the calculation method of polarization dimension information based on a sliding window is introduced, and various errors affecting the measurement are simulated and analyzed by using the established model. The results show that the size of the exposure angle will not have a great impact on the acquisition of polarization dimension information, and the influence of uneven rotation speed on the stocks parameter decreases with the increase of rotation speed. To ensure that the error of the Stocks parameter is less than 0.01, the absolute value of the initial deviation angle calibration error should be less than 0.573°, and the absolute value of the optical axis deflection angle should be less than 11.422°.

The laser radar using time-correlated single-photon counting technology has higher sensitivity, and the ability of target recognition can be increased by using polarization detection technology. A time-correlated single-photon counting imaging system based on polarization detection is built to image targets of different materials, the same depths and different depths. The results show that when the number of photons in a single pulse echo is 0.04, the polarization information of the target echo can be obtained by using the polarization detection method, and compared with the intensity image and the depth image, the edge of the target in the polarization image is clearer.

Images of different polarization state characteristics and intensities are collected for different camouflaged targets based on a four-channel polarized mid-wave infrared camera, according to the polarization equation, the polarization degree and polarization angle information of different targets are calculated, and the obvious polarization characteristics of the target are obtained by using a variety of image fusion algorithms. By comparing with the infrared intensity imaging of the target itself, it can be seen that the infrared polarization imaging can effectively identify the target using ordinary infrared camouflage through different image fusion algorithms. The research is very significant to the identification of camouflaged targets using infrared polarization.