Single-image super resolution via hashing classification and sparse representation

Single image super-resolution technology is a fundamental and important issue in the field of computer vision. This paper presents a hash-based classified dictionary learning method to reconstruct images. Firstly, calculate hash value of each image patches and classify the image patches according to its hash values. Then tightly sub-dictionary is learned for each cluster. For a given test image patch, corresponding sub-dictionary is selected by hash value, and then super-resolution reconstruction of this image is complete. The experimental results show that the proposed method is superior to the recently proposed dictionary learning methods for image super-resolution resolution in details and ensures quality of the reconstructed images.

[1]  Allen Y. Yang,et al.  Robust Face Recognition via Sparse Representation , 2009, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[2]  William T. Freeman,et al.  Learning Low-Level Vision , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[3]  D. Yeung,et al.  Super-resolution through neighbor embedding , 2004, CVPR 2004.

[4]  Moon Gi Kang,et al.  Super-resolution image reconstruction: a technical overview , 2003, IEEE Signal Process. Mag..

[5]  Peyman Milanfar,et al.  RAISR: Rapid and Accurate Image Super Resolution , 2016, IEEE Transactions on Computational Imaging.

[6]  Yücel Altunbasak,et al.  Color plane interpolation using alternating projections , 2002, 2002 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[7]  Owen Carmichael,et al.  Learning Low-level Vision Learning Low-level Vision , 2000 .

[8]  Harry Shum,et al.  A two-step approach to hallucinating faces: global parametric model and local nonparametric model , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[9]  D. Donoho.  For most large underdetermined systems of equations, the minimal 𝓁1‐norm near‐solution approximates the sparsest near‐solution , 2006 .