A Variational Bayesian Superresolution Approach Using Adaptive Image Prior Model

Figure 3

Results obtained by applying different approaches to LR car images corrupted with average blur and white Gaussian noise with SNR = 30 dB and their corresponding error images. Brighter pixels represent a large error in error image.

(a) Bicubic

(b) SAR

(c)

(d) TV

(e) BEP

(f) Ours

(g) Error image of bicubic interpolation

(h) Error image of SAR

(i) Error image of

(j) Error image of TV

(k) Error image of BEP

(l) Error image of ours