Research On Strong Background Interference Scattering Imaging

Due to the roughness of the surface of the scattering medium or the non-uniformity of the internal refractive index,random scattering events occur when light passes through the scattering media such as biological tissues,ground glasses,clouds,etc.,resulting in a speckle pattern.Aiming at the problem of imaging through scattering media,a series of methods have been developed in recent years,such as wavefront shaping,scattering holography,transmission matrix,optical phase conjugation,and speckle correlation.However,the imaging condition of these methods is basically limited in the laboratory environment,which does not consider the influence of the interference light in the ambient environment.For speckle correlation imaging,the presence of external background interference light is seriously reduced the contrast of speckle autocorrelation pattern,that makes its detailed structures submerged in the background interference noise and reducing the quality of reconstruction performance.Based on the characteristics of strong background interference light,this paper proposes a method for removing strong background interference light based on Zernike polynomial fitting and low-rank sparse matrix decomposition.It removes the influence of the interference light in a transform domain and analyzes different interference light conditions.We finally realize the imaging through the scattering media when the signalto-noise ratio is as low as-24 d B.The specific research content is as follows:(1)The characteristics of strong background interference light are studied,and it is found that the autocorrelation of strong background interference light is a combination of wavy background and spike noise.Based on this,the remove of a strong background interference light combining Zernike polynomial fitting and low-rank sparse matrix decomposition is proposed.The proposed method is removing the strong background interference light in two steps,and the removal performance in each step is verified by simulation.The strong background interference light removal model is successfully established,which paves the way for imaging through the scattering media under the strong background interference light.(2)In the laboratory environment,the data acquisition and target reconstruction through transmissive scattering media under strong background interference light are realized.Based on the established strong background interference light removal model,combined with the speckle correlation imaging method,we firstly simulate and verified the proposed method.Then we build the corresponding experimental setup to achieve the reconstruction through the scattering medium under a lower signal-to-noise ratio environment.Using this strategy,the anti-interference ability of the speckle correlation imaging method can be improved.(3)The imaging capabilities of the proposed method in this article are deeply researched in different scenarios.Firstly,the imaging limit of the traditional speckle correlation imaging method and the modified method is compared experimentally,and it is found that the achievable imaging signal-to-noise ratio of the traditional method is as low as-13 d B,while the method proposed in this paper can achieve a signal-to-noise ratio as low as-24 d B.Then,the imaging reconstruction of different targets under the interference of strong background light is realized;finally,the data is collected outdoor with a strong natural light,and the target reconstruction in the real natural scene is observed experimentally.In summary,we proposed a method for removing strong background interference light based on Zernike polynomial fitting and low-rank sparse matrix decomposition leading to the high performance of the scattering imaging.The research of the proposed method is expected to broaden the application scope of scattering imaging,promoting the development of scattering imaging technology from laboratory environment to natural environment applications,and to improve the possibility of scattering imaging technology used in daily life and industrial production.