TSS strict sensor model and its stable solution

High-resolution (HR) satellite images have been more and more applied to large-scale mapping; however, due to lack of RFM parameters or nature of high-order RPC model, the solutions occasionally fail. Based on a strict imaging model proposed by Weser, this paper, by adopting different functions to correct orbits and attitudes, analyze effects caused by sources of errors. Besides, because of invariant of the inner and exterior orientation elements within a strip of TSS and the change of attitudes, this paper adds virtual error equations to the solutions. Meanwhile, as TSS orientation parameters are highly correlated, we compare the results produced by spectrum-correlation, separated computation of linear-angle parameters, and Generalized Ridge Estimation.