Super-Resolution of Magnetic Resonance Images via Convex Optimization with Local and Global Prior Regularization and Spectrum Fitting

<div>Illustration of the Gerchberg algorithm. (a) The observed image (left) and its spectrum (right). (b) The interpretation of (a). The observed image is the sum of the true signal/spectrum (blue line) and the error signal/spectrum (red line). (c) The procedure followed by the algorithm. The error spectrum is reduced by iterating in both the signal and Fourier domains.</div>

International Journal of Biomedical Imaging

fig2

Figure 2

Figure 2: Super-Resolution of Magnetic Resonance Images via Convex Optimization with Local and Global Prior Regularization and Spectrum Fitting 