A5090588353- Advanced Image Processing Techniques
- Image Processing Techniques and Applications
- Image and Signal Denoising Methods
- Advanced Vision and Imaging
- Automated Road and Building Extraction
- Advanced Image and Video Retrieval Techniques
- Video Surveillance and Tracking Methods
Northwest University
2017-2019
Sparse representation has recently attracted enormous interests in the field of image super-resolution. The sparsity-based methods usually train a pair global dictionaries. However, only dictionaries cannot best sparsely represent different kinds patches, as it neglects two most important features: edge and direction. In this paper, we propose to novel pairs Direction Edge for For single-image super-resolution, training patches are, respectively, divided into clusters by new templates...
Image super-resolution based on learning dictionary has recently attracted enormous interests in the field of image super-resolution. In general, is trained from a large number training samples. this paper, we proposed to use nonlocal self-similarity filter blocks that are not similar structure. Based judgment method structural dissimilarity, can get small set dissimilar samples and then obtain pair high quality Structural Dissimilarity Learning Dictionary (SDLD). Extensive experiments (SR)...
Image super-resolution based on learning dictionary has recently attracted enormous interests. The learning-based methods usually train a pair of dictionaries from low-resolution and high-resolution image patches, ignoring the fact that patches have different structures. In this paper, we propose to set novel multi-pairs for categories which clustered by gaussian mixture model, instead global trained all patches. via patch prior guided clustering can express structure information well....
In this paper, we present an improved single image super-resolution method. The improvements are mainly attributed to block feature coding (BFC) that is select structurally dissimilar patches by the edge and direction features of patches. A structural dissimilarity learning dictionary (SDLD-BFC) pair trained on a small training set. Numerous experiments demonstrate efficient SDLD-BFC robust SDLDBFC Compared with other SR methods, significantly improves efficiency, while recovering good texture.