Imaging through scattering media is an important research topic in many fields, but it remains challenging. Imaging techniques based on optical memory effects have been extensively studied, but most methods require operations in dark environments. The extra interfering light submerges the speckle information in the spatial noise, resulting in a lower reconstruction quality of the image. In this paper, a principal component analysis method based on singular value decomposition is proposed for imaging under strong interference light. By selecting an appropriate singular value decomposition mode, the environment noise and effective speckle information are fitted to reconstruct the target object. The feasibility and reconstruction effect of this method are verified by experiments. The effective information extracted by the method proposed in this paper is very similar to that obtained under dark room conditions and strong interference light conditions. The selection of singular values is also discussed. This method can still extract speckle information about the unexposed part of the image even if some information is lost due to exposure. Our method can realize the separation of effective speckle information and ambient noise without prior knowledge, breaking through the limitation of the dark room environment of scattering imaging, and because our method can directly separate the target object information from the original speckle, it may be combined with other scattering imaging techniques.