Super-resolution reconstruction produces highresolution images from a set of low-resolution images of the same scene. In the last two and a half decades, many super-resolution algorithms have been proposed. These algorithms are usually very sensitive to their assumed models of data and noise, and also computationally expensive for many practical applications. In this paper we adopt computationally efficient prediction based sub-pel motion estimation to produce a fast super-resolution reconstruction that can also accommodate generic motion patterns. The proposed algorithm adaptively exploits the available high frequency content in adjacent video frames to generate high resolution video frames. Initial experiments showed promising results of around 2dB, PSNR improvement, over single frame bi-linear interpolation.