Wavelet-Based Image Interpolation Using Phase-Shift Compensation

Image interpolation in the wavelet domain usually takes advantage of probabilistic models of intra-scale statistics and inter-scale dependency. Parameter estimation for the probabilistic model of wavelet coefficients is essential to the wavelet-domain interpolation. However, phase uncertainty, one well-known drawback of the orthogonal wavelet transforms, makes it difficult to estimate parameters. In this paper, we adopt a phase shifting matrix in order to improve the accuracy of estimation. Nonlinear modeling to capture the inter-scale characteristics is also described. The experimental results show that the proposed method outperforms the previous wavelet-domain interpolation method as well as the conventional bicubic method