Super-resolution algorithm through neighbor embedding with new feature selection and example training

An improved super-resolution algorithm through neighbor embedding with new feature selection and example training is proposed for single image super resolution reconstruction. Firstly, we take the DCT coefficients as the feature vectors, and then adaptively choose neighbors by k-means clustering algorithm. Finally, we learn the neighborhood relationship between interpolated image from low resolution image and its corresponding high resolution image. The experimental results show that the improved algorithm can not only achieve a better recovery of a single low resolution image comparing with the original neighbor embedding algorithm, but also reduce the computational complexity.