Image Deblurring Using Adaptive Sparse Domain Selection and Adaptive Regularization

D. Suresh

Abstract


Problem Statement: Image restoration (IR) aims to reconstruct a high-quality image from its degraded measurement. Approach: As a powerful statistical image modeling technique, sparse representation has been successfully used in various image restoration applications. The contents of images can vary significantly across different images or different patches in a single image, and then, for each selected patch, one set of bases are adaptively selected to characterize the local sparse domain. Then two adaptive regularization terms are used into the sparse representation framework. First, a set of autoregressive (AR) models are adaptively selected to regularize the image local structures. Second, the image non-local self-similarity is introduced as another regularization term estimated for better image restoration performance. Conclusion: Iterative experiments using adaptive sparse domain selection and adaptive regularization, the proposed method achieves much better results than many state-of-the-art algorithms in terms of both PSNR and visual perception.

Key Words: Sparse representation, image restoration, deblurring, super-resolution, regularization.


Full Text:

PDF


DOI: https://doi.org/10.26483/ijarcs.v3i3.1115

Refbacks

  • There are currently no refbacks.




Copyright (c) 2016 International Journal of Advanced Research in Computer Science