With the continuous development of marine resources and the increasingly prominent issues of territorial sea,topics such as underwater target detection,underwater environmental monitoring and maritime military reconnaissance have received more extensive attention.However,due to the absorption and scattering of light by the medium in the water,the obtained images have problems such as low contrast,low definition,loss of detail and color distortion,which reduce the application value of the images.Hyperspectral polarization imaging technology can acquire the spatial information,spectral information and polarization information of the targets simultaneously.The polarization information at each spatial position of the target can be obtained with nanometer-level spectral resolution,providing a richer information source for target detection.Compared with traditional intensity imaging technology,hyperspectral polarization imaging technology is more effective for underwater target detection and recognition.Aiming at the task of improving the detection capability of the imaging system and recovering the underwater images,this paper proposes a hyperspectral polarization imaging technology based on liquid crystal modulation,which uses a combination of liquid crystal tunable filter splitting and ferroelectric liquid crystal high-speed polarization modulation to achieve the synchronous detection of target wide spectrum and full polarization information.The system has the characteristics of compact structure,no moving parts,high stability and staring imaging.In this paper,the filtering principle and process of the liquid crystal tunable filter are simulated and verified,and the fast axis angles of the ferroelectric liquid crystals and the wave plates of the polarization modulation module are optimized by using the genetic algorithm to reduce the system noise and error propagation.Meanwhile,this paper carries out the research on image recovery in the environment of underwater scattering media,through indepth analysis of the current algorithms,a novel underwater image recovery algorithm based on spectral polarization imaging are proposed in view of the shortcomings of the current algorithms,and the proposed recovery algorithm combines the respective advantages of the Stokes vector-based polarization recovery method and the multispectral fusion method to further improve the clarity and imaging distance of underwater imaging.By building an experimental platform to conduct indoor and outdoor hyperspectral polarization imaging experiments respectively,and obtain full polarization information under all spectrum bands of the target.The experimental results can accurately reflect the spectral polarization characteristics of the target,indicating that the system has good measurement performance.In addition,the underwater hyperspectral polarization imaging experiment verifies the effectiveness of the underwater image recovery method based on spectral polarization imaging,and the quantitative comparison and analysis of the experimental results processed by the proposed method and the traditional methods show that the proposed method can obtain more detailed information of the target and has a wider application range.The hyperspectral polarization imaging system based on liquid crystal modulation proposed in this paper combined with the underwater image recovery method based on spectral polarization imaging provides an effective technical solution for underwater target detection and has potential application value in underwater real-time clear imaging. |