According to the error sources and their error amounts obtained by the remote sensing imaging coupled typical error sources inversion method, we can improve the imaging quality of optical systems and make high-quality remote sensing images more useful in military and civilian fields. Based on the distorted remote sensing images, this paper proposes a remote sensing imaging coupled typical error sources inversion method, which can accurately invert the typical error sources of remote sensing imaging and their error amounts. Firstly, a set of coupled typical error decoupled equations are constructed according to the modulation transfer function model of typical error sources and the decoupled principle of the coupled error sources. The initial values of coupled typical error sources are subsequently determined based on the Deep Residual Shrinkage Network (DRSN). Finally, the Levenberg Marquardt-Particle Swarm Optimization (LM-PSO) hybrid optimization algorithm is used to solve the system of coupled typical error decoupled equations to invert the typical error sources and their error amounts of the remote sensing imaging system. The experimental results show that the relative error between the inverse value and the real value of the coupled typical error sources of the distorted remote sensing images by the method in this paper does not exceed 20% at most, and most of them are below 10%, which has excellent inversion performance.