Specification of the observation model for regularized image up-sampling

Regularization is one of the most promising methods for image up-sampling, which is an ill-posed inverse problem. A key element of such a regularization approach is the observation model relating the observed lower resolution (LR) image to the desired higher resolution (HR) up-sampled image, used in the data-fidelity term of the regularization cost function. This paper presents an algorithm to determine this observation model based on a model of the physical acquisition process for the LR image, and the ideal acquisition process for the desired HR image, both from the same underlying continuous image. The method is illustrated with typical scenarios corresponding to LR and HR cameras modeled by either Gaussian or rectangular apertures. Experiments with some regularized image up-samplers demonstrate the importance of using the correct, adapted observation model as determined by our algorithm.

[1]  E. Dubois,et al.  The sampling and reconstruction of time-varying imagery with application in video systems , 1985, Proceedings of the IEEE.

[2]  Giovanni Ramponi,et al.  A simple edge-sensitive image interpolation filter , 1996, Proceedings of 3rd IEEE International Conference on Image Processing.

[3]  Eric Dubois,et al.  Regularized image up-sampling using a new observation model and the level set method , 2003, Proceedings 2003 International Conference on Image Processing (Cat. No.03CH37429).

[4]  Michael Unser,et al.  Splines: a perfect fit for signal and image processing , 1999, IEEE Signal Process. Mag..

[5]  Ronald N. Bracewell,et al.  The Fourier Transform and Its Applications , 1966 .

[6]  Michael Elad,et al.  Restoration of a single superresolution image from several blurred, noisy, and undersampled measured images , 1997, IEEE Trans. Image Process..

[7]  Aggelos K. Katsaggelos,et al.  Resolution enhancement of monochrome and color video using motion compensation , 2001, IEEE Trans. Image Process..

[8]  E. Dubois,et al.  Linear phase and symmetries for multidimensional FIR filters over lattices , 1998 .

[9]  Luigi Cinque,et al.  Improvements to image magnification , 2002, Pattern Recognit..

[10]  Takeo Kanade,et al.  Limits on super-resolution and how to break them , 2000, Proceedings IEEE Conference on Computer Vision and Pattern Recognition. CVPR 2000 (Cat. No.PR00662).

[11]  Otto H. Schade.  Image Reproduction by a Line Raster Process , 1973 .

[12]  Jan P. Allebach.  7.1 – Image Scanning, Sampling, and Interpolation , 2005 .

[13]  Robert L. Stevenson,et al.  Super-resolution from image sequences-a review , 1998, 1998 Midwest Symposium on Circuits and Systems (Cat. No. 98CB36268).

[14]  Robert L. Stevenson,et al.  A Bayesian approach to image expansion for improved definitio , 1994, IEEE Trans. Image Process..

[15]  Anil K. Jain.  Fundamentals of Digital Image Processing , 2018, Control of Color Imaging Systems.

[16]  Eric Dubois,et al.  Crafting the observation model for regularized image up-sampling , 2003, 2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03)..

[17]  Rabab Kreidieh Ward,et al.  A new edge-directed image expansion scheme , 2001, Proceedings 2001 International Conference on Image Processing (Cat. No.01CH37205).

[18]  P. Welch.  The use of fast Fourier transform for the estimation of power spectra: A method based on time averaging over short, modified periodograms , 1967 .

[19]  Deepu Rajan,et al.  Generation of super-resolution images from blurred observations using Markov random fields , 2001, 2001 IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings (Cat. No.01CH37221).

[20]  Eric Dubois,et al.  Design of optimal camera apertures adapted to display devices over arbitrary sampling lattices , 2004, IEEE Signal Processing Letters.