The Unmanned Aircraft Systems (UAS) have been widely applied for reconnaissance and surveillance by exploiting the information collected from the digital imaging payload. However, the analysis of real surveillance video is limited from time to time because of the constraints of imaging equipment, winds buffeting the vehicle, less than ideal atmospheric conditions, etc. As a consequence, the superresolution (SR) on low-resolution (LR) UAS surveillance video frames, becomes a critical requirement for UAS video processing and an important pre-step for further effective image understanding. In this paper we develop a novel robust super-resolution framework which does not require the construction of sparse matrices. The proposed method implements image operations in the spatial domain and applies an iterative process to construct super-resolution images from the overlapping UAS surveillance video frames. Our method adopts the sum of forward-projection fitting errors (from SR images to LR images) and backward-projection fitting errors (from LR images to SR images) as the optimization measure. The proposed algorithm has been tested on two sets of real UAS video data and its performance is compared with those from other SR algorithms. The experimental outcomes show the good performance of the proposed algorithm in comparison with other algorithms.