Research Of Underwater Image Recovery Based On Mueller Matrix And Stokes Vector

As one of the basic physical properties of light waves,polarization has significant advantages in the application areas of light intensity and spectral inefficiency due to its characteristics of "weak light intensification and strong light weakening",light intensity insensitivity and multi-dimensionality.Especially in complex scattering environments such as turbid underwater and haze,polarization imaging demonstrates superior optical information extraction,recognition and processing capabilities.As typical polarization characterization parameters,Stokes vectors and Mueller matrices are commonly used to jointly characterize the polarization properties of light and matter as well as the changes of polarization states before and after light-matter interactions(reflection,transmission,etc.).The main work of this paper is as follows.1.Based on the classical underwater scattering polarization imaging model,a Mueller matrix-based scattering polarization imaging system under water is proposed.By calculating the Mueller matrix in the background region,the polarization state of the polarization generating device is adjusted so that the polarization of the backscattered light is maximized;the polarization analysis device is used to effectively suppress the polarization component in the backscattered light,and a better effect of suppressing the scattered light is achieved.The method is validated by experiments under different conditions and compared with the conventional underwater polarization imaging method by quantifying the image enhancement(EME)and standard deviation(STD)of the data.The results all show that the method has a better effect of suppressing scattered light and a higher image recovery,which can effectively improve the underwater imaging quality and visual distance.2.A three-degree-of-freedom polarization differential imaging algorithm is proposed to address the problems of low degrees of freedom and limited image quality enhancement of traditional polarization differential imaging.The degree of freedom of polarization differential imaging is increased by introducing correction parameters without limiting the azimuthal range of the polarizer.This paper also compares with two classical polarization differential imaging algorithms by designing experiments in different scenes,and the method achieves a Michelson contrast ratio(Contrast ratio)of0.851 and a corresponding enhancement ratio of up to 170.2%.The results show that the image recovered by the three-degree-of-freedom polarization differential method has low distortion and high definition,which can further improve the effect of polarization differential imaging on scattering imaging.3.Using Lab VIEW software,this paper designs a polarization imaging software platform based on two different types of polarization cameras(split-focus plane and split-amplitude)respectively.The three-degree-of-freedom polarization differential algorithm is integrated into the software system to realize the functions of automatic camera control,real-time acquisition and display of polarization information,and simultaneous display of backscattering imaging results.Meanwhile,a module is designed to read polarization images from local files and realize the descattered imaging for the demand of detached camera use in practical applications.The combination of the de-scattering imaging algorithm with practical applications is of great significance for the development and practicalization of underwater polarimetric imaging.