Classification of methods in increasing image resolution over several frames

Authors: Sirotkina P.Yu.
Published in issue: #5(94)/2024
DOI:
Category: Informatics, Computer Engineering and Control | Chapter: Information Technology. Computer techologies. Theory of computers and systems
Keywords: digital image, signal processing, super-resolution, convolutional neural networks, interpolation, reconstruction, regularization, set theory, patches
Published: 18.11.2024

The paper is devoted to reviewing the methods in increasing the image resolution (the so-called, super-resolution problem) based on several frames of the same object. It considers in brief the subject area and main approaches to solving the problem. They include interpolation-based methods, reconstruction methods with regularization, methods based on the set theory, example-oriented methods (patch extraction), and methods for optimizing the convolutional neural networks. The paper provides a critical analysis of methods increasing image resolution based on several frames with identifying the areas of most efficient use based on the following criteria: computational complexity, processing quality, need for post- or pre-processing. The paper concludes on the applicability area for the considered methods. To achieve the best processing quality, it is advisable to introduce the methods based on the convolutional neural networks.

References

[1] Nasonov A.V., Krylov A.S. Fast super-resolution of images using weighted median filtering. Digital signal processing and its application. 12th International Conference and exhibition: sat. tr. Moscow, Lomonosov Moscow State University Publ., 2010, vol. 2, pp. 101–104. (In Russ.).

[2] Park S.C., Park M.K., Kang M.G. Super-resolution image reconstruction: a technical overview. IEEE signal processing magazine, 2003, vol. 20, no. 3, pp. 21–36. https://doi.org/10.1109/MSP.2003.1203207

[3] Kokoshkin A.V. et al. Estimation of errors in image synthesis with super resolution based on the use of multiple frames. Computer Optics, 2017, vol. 41, No. 5, pp. 701–711. https://dx.doi.org/10.18287/2412-6179-2017-41-5-701-711

[4] Wang Q., Tang X., Shum H. Patch based blind image super resolution. Tenth IEEE International Conference on Computer Vision, IEEE, 2005, vol. 1, pp. 709–716. https://doi.org/10.1109/ICCV.2005.186

[5] Freeman W.T., Jones T.R., Pasztor E.C. Example-based super-resolution. IEEE Computer graphics and Applications, 2002, vol. 22, no. 2, pp. 56–65.

[6] Katsaggelos A.K. Digital image restoration. Springer Publishing Company, Incorporated, 2012.

[7] Fan C. et al. POCS Super-resolution sequence image reconstruction based on improvement approach of Keren registration method. Sixth International Conference on Intelligent Systems Design and Applications, IEEE, 2006, vol. 2, pp. 333–337. https://doi.org/10.1109/ISDA.2006.253857

[8] Stark H., Oskoui P. High-resolution image recovery from image-plane arrays, using convex projections. JOSA A, 1989, vol. 6, no. 11, pp. 1715–1726. https://doi.org/10.1364/josaa.6.001715

[9] Bredikhin A.I. Learning algorithms for convolutional neural networks. Bulletin of Yugra State University, 2019, No. 1 (52), pp. 41–54. (In Russ.).

[10] Umehara K., Ota J., Ishida T. Application of super-resolution convolutional neural network for enhancing image resolution in chest CT. Journal of digital imaging, 2018, vol. 31, pp. 441–450. https://doi.org/10.1007/s10278-017-0033-z

[11] Dong C. et al. Learning a deep convolutional network for image super-resolution. Computer Vision-ECCV 2014: 13th European Conference. Proceedings, Zurich, Switzerland, Springer International Publishing, 2014, part IV, pp. 184–199. https://doi.org/10.1007/978-3-319-10593-2_13

[12] Senov A. Projective approximation based quasi-Newton methods. Proc. of International Workshop on Machine Learning, Optimization, and Big Data. Springer, Cham, 2017, pp. 29–40. (In Russ.). https://doi.org/10.1007/978-3-319-72926-8_3

[13] Senov A.A. Deep learning in the problem of image superresolution reconstruction. Stochastic Optimization in Computer Science, 2017, vol. 13, No. 2, pp. 38–57. (In Russ.).