Дослідження та аналіз методів забезпечення надвисокої роздільної здатності зображень на основі машинного навчання

2017;
: pp. 68-76
Accepted: March 28, 2017
Authors:
1
Lviv Politechnik National University, Department of Publishing Information Technologies

In this article the methods of image superresolution based on machine learning are
investigated. The work of different groups of these methods are analyzed. Basic features of this
methods are describing. On the basis of practical experiments comparative analysis (by the
criterion PSNR) of the superresolution methods in the case of one input image from different
classes were conducted. Experimentally found that the best results are obtained in case of
using the method based on the convolutional neural network. Despite the requirement on the
time and resources which are needed to implement the training procedures of this method, its
training model can be used in the processing of images of different classes.

1. Rasrollahi K. Super-resolution: a comprehensive survey / K. Nasrollahi, T. B. Moeslund //
Machine Vision and Applications. – 2014. – Vol. 25, № 6. – P. 1423–1468.

2. Xiong Z. Robust web imagevideo
super-resolution / Z. Xiong, X. Sun, and F. Wu // IEEE Transactions on Image Processing. – 2010. –
Vol. 19, № 9. – P. 2017–2028.

3. Sun J. Image hallucination with primal sketch priors / J. Sun,
N. N. Zheng, H. Tao, and H. Y. Shum // Computer Vision and Pattern Recognition: proc. of intern. conf.,
Madison, Wisconsin, 18–20 June 2003. – IEEE: Computer Society, 2003. – Vol. 2. – P. 729–736.

4. Srivastava A. On advances in statistical modeling of natural images / A. Srivastava, A. B. Lee,
E. P. Simoncelli, S.–C. Zhu // Journal of Mathematical Imaging and Vision. – 2003. – Vol. 18, № 1. –
P. 17–33.

5 Ting Li Image Super-Resolution using sharpened gradient profile prior: thesis…. master of
science: electrical engineering / Li, Ting. – Dallas, 2012. – 65 p.

6. Sun J. Gradient profile prior and its
applications in image super–resolution and enhancement / J. Sun, J. Sun, Z. Xu, H. Y. Shum // IEEE
Transactions on Circuits and Systems for Video Technology. – 2011. – Vol. 20, № 6. – P. 1529–1542.

7. Патент US 20100086227 (A1), США Image super–resolution using gradient profile prior / Jian Sun,
Heung–Yeung Shum; Microsoft Corporation. – US 12/245,712; 04.10.2008; 08.08.2010.

8. Казакова Н. Ф.
Синтез методу виділення контурів у системах ідентифікації на основі усереднення перепадів
76
яскравості / Н. Ф. Казакова, О. О. Фразе–Фразенко // Інформаційна безпека. – 2013. – № 2. – С. 48–57.

9. Yu L. Robust Single Image Super–resolution based on Gradient Enhancement / Licheng Yu, Hongteng
Xu, Yi Xu, Xiaokang Yang // Signal and Information Processing: annual summit and conf., Hollywood, CA,
3–6 Dec. 2012. – Asia–Pacific Signal and Information Processing Association, 2008. –
P. 1 – 6.

10. He He Single Image Super–Resolution using Gaussian Process Regression / He He, Wan–Chi
Siu // Computer Vision and Pattern Recognition: proc. of intern. conf., Colorado, USA, 20–24 June
2011. – IEEE: Computer Society, 2011. –P. 449–456.

11. E. Mjolsness Neural networks, pattern
recognition, and fingerprint hallucination: thesis … doctor of philosophy: 5198: TR:85 / Eric Mjolsness. –
California, 1985. – 79 p.

12. Datsenko D. Example–based single document image super–resolution:
A global map approach with outlier rejection / D. Datsenko, M. Elad // Journal of Multidimensional
Systems and Signal Processing. – 2007. – № 2. – P. 103 – 121.

13. Liu C. Face Hallucination: Theory and
Practice / Ce Liu, Heung–Yeung Shum, William T. Freeman // International Journal of Computer Vision. –
2007. – Vol. 75, № 1. – P. 115 – 134.

14. Sua C. Steerable pyramid–based face hallucination / Congyong
Sua, Yueting Zhuanga, Li Huanga, Fei Wua // Pattern Recognition. – 2005. – Vol. 38, № 6. – P. 813 – 824.

15. J. Yang, J. Wright, T. Huang, and Y. Ma, “Image super-resolution as sparse representation of raw
image patches,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Jun. 2008,
pp. 1–8.

16. R. Zeyde, M. Elad, and M. Protter, “On single image scale-up using sparserepresentations,”
in Curves and Surfaces. Springer, 2010.

17. Freeman T. William Learning low–level vision / William
T. Freeman, Egon C. Pasztor, Owen T. Carmichael // International Journal of Computer Vision. – 2000. –
Vol. 40, № 1. – P. 25 – 47.

18. Gajjar Prakash Zoom Based Super–Resolution: A Fast Approach Using
Particle Swarm Optimization / Prakash Gajjar, Manjunath Joshi // Image and Signal Processing, Lecture
Notes in Computer Science. – 2010. – Vol. 6134. – P. 63 – 70.

19. E. Mjolsness Neural networks, pattern
recognition, and fingerprint hallucination: thesis … doctor of philosophy: 5198: TR:85 / Eric Mjolsness. –
California, 1985. – 79 p.

20. Nguyen Q. M. Superresolution mapping using a Hopfield neural network with
LIDAR data / Minh Quang Nguyen, Peter M. Atkinson, Hugh G. Lewis // Geoscience and Remote Sensing
Letters. – 2005. – Vol. 2, № 3. – P. 366–370.

21. Nguyen Q. M. Superresolution Mapping Using a Hopfield
Neural Network With Fused Images / Minh Quang Nguyen, Peter M. Atkinson, Hugh G. Lewis // IEEE
Transaction on geoscience and remote rensing. – 2002. – Vol. 40, № 3. – P. 736–749.

22. Thornton M. W.
A linearised pixel–swapping method for mapping rural linear land cover features from fine spatial
resolution remotely sensed imagery / M. W. Thornton, P. M. Atkinson, and D. A. Holland // Computers &
Geosciences. – 2007. – Vol. 33, № 10. – P. 1261–1272.

23. Pan F. New image super–resolution scheme
based on residual error restoration by neural networks. / F. Pan, L. Zhang // Optical Engineering. –
2003. – Vol. 42, № 10 – P. 3038–3046.

24. Dong C. Image Super–Resolution Using Deep Convolutional
Networks / Dong Chao, Loy Chen Change, He Kaiming, Tang Xiaoou // IEEE Transactions on Pattern
Analysis and Machine Intelligence, Preprint. – 2015. – P. 14.

25. Ізонін І.В. Нейромережевий метод
зміни роздільної здатності зображень / І. В. Ізонін, Р. О. Ткаченко, Д. Д. Пелешко, Д. А. Батюк //
Системи обробки інформації. – 2015. – Вип. 9(134). – C. 30–34.

26. Izonin I. Learning-based image
super-resolution using weight coefficients of synaptic connections / Ivan Izonin, Roman Tkachenko,
Dmytro Peleshko, Taras Rak, Danylo Batyuk // Computer science and information technologies: proc. of X
intern. scien. and techn. conf., 14–17 Sep. 2015 y., Lviv, Ukraine. – Lviv: Lviv Polytechnic Publishing
House, 2015. – P. 25–29.