Limits on Super-Resolution and How to Break Them
IEEE Transactions on Pattern Analysis and Machine Intelligence
Deterministic edge-preserving regularization in computed imaging
IEEE Transactions on Image Processing
Image quality assessment: from error visibility to structural similarity
IEEE Transactions on Image Processing
A segmentation-based regularization term for image deconvolution
IEEE Transactions on Image Processing
Region-Based Super Resolution for Video Sequences Considering Registration Error
PSIVT '09 Proceedings of the 3rd Pacific Rim Symposium on Advances in Image and Video Technology
Kurtosis-based super-resolution algorithm
ICME'09 Proceedings of the 2009 IEEE international conference on Multimedia and Expo
Region-based weighted-norm with adaptive regularization for resolution enhancement
Digital Signal Processing
Hi-index | 0.00 |
In this paper a novel high-order statistics (HOS) based regularized algorithm for image super-resolution reconstruction is proposed. In this method, the image is divided into various regions according to the local forth order statistics. The segmentation label is then used to determine the weighted operator of the regularization term. In this way, different regularization terms are applied depending on local characteristics and structures of the image. The proposed image achieves anisotropic diffusion for edge pixels and isotropic diffusion for flat pixels. Experimental results demonstrate that the proposed method performs better than the conventional methods and has high PSNR and MSSIM with sharper edges.