Name
Abstract | ||
|---|---|---|
Block based compressed sensing (BCS) is a novel framework in image signals sampling and recovery due to its advantages in terms of both low sampling burden and lightweight recovery complexity. In this paper, we propose a novel asymmetric BCS scheme to further improve the image recovery accuracy. In the sampling process, image blocks are partitioned into smaller sub-blocks, and those small sub-blocks are used to allocate sampling resources. In the recovery process, the small sub-blocks with similar feature information are assembled into virtual blocks with larger size, and the corresponding transforming coefficients are then more compressible. The proposed scheme improves the recovered images from the fairer resources allocation and much greater compressibility. The experimental results demonstrate that, compared to the existing BCS approaches, our proposed scheme has higher recovery quality, without increasing sampling and recovery complexity. |
Year | Venue | Field |
|---|---|---|
2018 | ICME | Resource management,Iterative reconstruction,Sampling process,Computer vision,Pattern recognition,Computer science,Image coding,Sampling (statistics),Artificial intelligence,Image recovery,Cluster analysis,Compressed sensing |
DocType | Citations | PageRank |
Conference | 0 | 0.34 |
References | Authors | |
0 | 4 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Siwang Zhou | 1 | 4 | 2.39 |
| Shuzhen Xiang | 2 | 0 | 0.68 |
| Xingting Liu | 3 | 0 | 0.68 |
| Heng Li | 4 | 21 | 7.48 |