Image interpolation using shearlet based iterative refinement

• We develop an image interpolation algorithm exploiting sparse representations for natural images.
• For sparse modeling, a shearlet dictionary is chosen to yield a multiscale directional representation.
• We start with an initial estimate of high resolution image and iteratively refine it towards an improved solution.
• Objective and subjective comparison to many well known methods is provided over a dataset of 200 images.

This paper proposes an image interpolation algorithm exploiting sparse representation for natural images. It involves three main steps: (a) obtaining an initial estimate of the high resolution image using FIR filtering, (b) promoting sparsity in a selected dictionary through hard thresholding to obtain an approximation, and (c) extracting high frequency information from the approximation to refine the initial estimate. For the sparse modeling, a shearlet dictionary is chosen to yield a multiscale directional representation. The proposed algorithm is compared to several state-of-the-art methods to assess its objective and subjective performance. Compared to the cubic spline interpolation method, an average PSNR gain of around 0.8 dB is observed over a dataset of 200 images.