Abstract
A novel approach for super-resolution using kernel regression technique is presented in this paper. The new algorithm uses several low resolution video frames to estimate unknown pixels in a high resolution frame using kernel regression employing adaptive Gaussian kernel. Experiments conducted on several video streams to evaluate the effect of the proposed algorithm showed improved performance when compared with other state of the art techniques. This resolution enhancement technique is simple and easy to implement and it can be used as a software alternative to obtain high quality and high resolution video streams from low resolution versions.