Realizing the Effective Detection of Tumor in Magnetic Resonance Imaging using Cluster-Sparse Assisted Super-Resolution

Recently, significant research has been done in Super-Resolution (SR) methods for augmenting the spatial resolution of Magnetic Resonance (MR) images, which aids physician improved disease diagnoses. Single SR have drawbacks; they fail to capture self-similarity non-local patches and are not robust noise. To exploit intrinsic sparsity MR this paper proposes use Cluster-Sparse Assisted Super-Resolution. This method effectively captures similarity non-locally positioned by training on clusters using a self-adaptive dictionary. also leads better edge texture detection. Experiments show that brain images results enhanced detection lesions leading diagnosis.