Target recovery through scattering media is an important aspect of optical imaging. Although various algorithms combining deep-learning methods for target recovery through scattering media exist, they have limitations in terms of robustness and generalization. To address these issues, this study proposes a data-decoupled scattering imaging method based on autocorrelation enhancement. This method constructs basic-element datasets, acquires the speckle images corresponding to these elements, and trains a deep-learning model using the autocorrelation images generated from the elements using speckle autocorrelation as prior physical knowledge to achieve the scattering recovery imaging of targets across data domains. To remove noise terms and enhance the signal-to-noise ratio, a deep-learning model based on the encoder–decoder structure was used to recover a speckle autocorrelation image with a high signal-to-noise ratio. Finally, clarity reconstruction of the target is achieved by applying the traditional phase-recovery algorithm. The results demonstrate that this process improves the peak signal-to-noise ratio of the data from 15 to 37.28 dB and the structural similarity from 0.38 to 0.99, allowing a clear target image to be reconstructed. Meanwhile, supplementary experiments on the robustness and generalization of the method were conducted, and the results prove that it performs well on frosted glass plates with different scattering characteristics.

Due to the effect of the complex scattering medium, the photons carrying target information will be attenuated when passing through scattering media, and target localization is difficult. The resolution of the target-position information from scattered images is crucial for achieving accurate target localization in environments such as dense fog in military applications. In this paper, a target localization network incorporating an attention mechanism was designed based on the robust feature resolution ability of neural networks and the characteristics of scattering formation. A training dataset with basic elements was constructed to achieve data decoupling, and then realize the position estimation of targets in different domains in complex scattering environments. Experimental validation showed that the target was accurately localized in speckle images with different domain data by the above method. The results will provide ideas for future research on the localization of typical targets in natural scattering environments.