Because they are affected by imaging conditions, aliasing, noise, etc, imaging systems are unable to obtain all of the information contained in an original scene. Super-resolution (SR) reconstruction is important for the application of image data to increase the resolution of images. In this article, an example-based algorithm is proposed to implement SR reconstruction by single-image. The mapping function between low-resolution (LR) and high-resolution (HR) images is learned by using the method of regularized regression. Then, finding the optimal sparse subset of the training data set by kernel matching pursuit (KMP). The results show that this method can recover detailed information of images, and the computational cost is reduced compared to other example-based SR methods.
|