Skip to main content
SpringerLink
Log in

Image adaptive sampling using reinforcement learning

  • Published:
Multimedia Tools and Applications Aims and scope Submit manuscript

Abstract

Adaptive sampling and mesh representation of images play an important role in image compression and vectorization. In this paper, a multi-points stochastic gradient multi-armed bandits algorithm, a generalization of the gradient bandit algorithm, is presented to adaptively sample points in images. By modeling the adaptive image sampling as a multi-arm selection decision-making problem, we first propose an efficient action selection strategy based on a parameterized probability distribution, and then define an adaptive reward function according to the restored image of Delaunay triangulation and a feature map function, and the reward function can overcome the sparse reward issue effectively. As a result, the proposed multi-points stochastic gradient multi-armed bandits algorithm is used to evaluate the reward of each action. At last, a prescribed number of sampling points are selected using a simple and effective strategy according to the average reward of each pixel. The quality of reconstructed images based on sampled points is estimated, and experimental results demonstrate the proposed algorithm achieves a better reconstruction accuracy than that of existing methods.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Algorithm 1
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7

Similar content being viewed by others

Data Availability

Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.

References

  1. Abramenko O, Jung A (2019) Graph signal sampling via reinforcement learning. In: ICASSP 2019-2019 IEEE International Conference on Acoustics Speech and Signal Processing (ICASSP). IEEE, pp 3077–3081

  2. Adams MD (2008) An efficient progressive coding method for arbitrarily-sampled image data. IEEE Signal Process Lett 15:629–632

    Google Scholar 

  3. Adams MD (2011) A flexible content-adaptive mesh-generation strategy for image representation. IEEE Trans on Image Process 20(9):2414–2427

    MathSciNet  Google Scholar 

  4. Ahmed AG, Guo J, Yan M, Franceschia JY, Zhang X, Deussen O (2016) A simple push-pull algorithm for blue-noise sampling. IEEE Trans Vis Comput Graphics. 23(12):2496–2508

    Google Scholar 

  5. Badrinarayanan V, Kendall A, Cipolla R (2017) Segnet: a deep convolutional encoder-decoder architecture for image segmentation. IEEE Trans Pattern Anal Mach Intell 39(12):2481–2495

    Google Scholar 

  6. Barr I (2017) Images to triangles. https://github.com/ijmbarr/images-to-triangles

  7. Battiato S, Gallo G, Messina G (2004) SVG rendering of real images using data dependent triangulation. In: Proceedings of the 20th spring conference on Computer graphics, pp 185–192

  8. Brankov JG, Yang Y, Galatsanos NP (2003) Image restoration using content-adaptive mesh modeling. In: Proceedings 2003 International conference on image processing (Cat. No. 03CH37429). vol 2. IEEE, pp 997–1000

  9. Brankov JG, Yang Y, Wernick MN (2004) Tomographic image reconstruction based on a content-adaptive mesh model. IEEE Trans Med Imaging 23 (2):202–212

    Google Scholar 

  10. Chowdhary CL, Patel PV, Kathrotia KJ, Attique M, Perumal K, Ijaz MF (2020) Analytical study of hybrid techniques for image encryption and decryption. Sensors 20(18):5162

    Google Scholar 

  11. Dai Q, Chopp H, Pouyet E, Cossairt O, Walton M, Katsaggelos AK (2019) Adaptive image sampling using deep learning and its application on X-ray fluorescence image reconstruction. IEEE Trans Multimed 22(10):2564–2578

    Google Scholar 

  12. De Goes F, Breeden K, Ostromoukhov V, Desbrun M (2012) Blue noise through optimal transport. ACM Transactions on Graphics (TOG) 31 (6):1–11

    Google Scholar 

  13. Eldar Y, Lindenbaum M, Porat M, Zeevi YY (1997) The farthest point strategy for progressive image sampling. IEEE Trans Image Process 6 (9):1305–1315

    Google Scholar 

  14. Fakhari A, Kiani K (2021) A new restricted boltzmann machine training algorithm for image restoration. Multimed Tools Appl 80(2):2047–2062

    Google Scholar 

  15. Fattal R (2011) Blue-noise point sampling using kernel density model. ACM Transactions on Graphics (TOG) 30(4):1–12

    Google Scholar 

  16. Felzenszwalb PF, Huttenlocher DP (2004) Efficient graph-based image segmentation. Int J Comput Vis 59(2):167–181

    Google Scholar 

  17. Floyd RW (1976) An adaptive algorithm for spatial gray-scale. In: Proc Soc Inf Disp, vol 17, pp 75–77

  18. García MA, Vintimilla BX, Sappa AD (1999) Efficient approximation of gray-scale images through bounded error triangular meshes. In: Proceedings 1999 International conference on image processing (Cat. 99CH36348). vol 1. IEEE, pp 168–170

  19. Gevers T, Smeulders A (1997) Combining region splitting and edge detection through guided Delaunay image subdivision. In: Proceedings of IEEE Computer society conference on computer vision and pattern recognition. IEEE, pp 1021–1026

  20. Grogan S (2016) Body image: Understanding body dissatisfaction in men, women and children Taylor & Francis

  21. Gu K, Liu H, Xia Z, Qiao J, Lin W, Thalmann D (2021) PM2.5 Monitoring: Use information abundance measurement and wide and deep learning. IEEE Trans Neural Netw Learn Syst 32(10):4278–4290

    Google Scholar 

  22. Gu K, Xia Z, Qiao J (2019) Stacked selective ensemble for PM 2.5 forecast. IEEE Trans Instrument Meas 69(3):660–671

    Google Scholar 

  23. Gu K, Xia Z, Qiao J, Lin W (2019) Deep dual-channel neural network for image-based smoke detection. IEEE Trans Multimed 22(2):311–323

    Google Scholar 

  24. Gu K, Zhang Y, Qiao J (2020) Ensemble meta-learning for few-shot soot density recognition. IEEE Trans Indus Inf 17(3):2261–2270

    Google Scholar 

  25. Haralick RM, Shapiro LG (1985) Image segmentation techniques. Comput Vis Graphics Image Process 29(1):100–132

    Google Scholar 

  26. He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 770–778

  27. Hussain R, Karbhari Y, Ijaz MF, Woźniak M, Singh PK, Sarkar R (2021) Revise-net: Exploiting reverse attention mechanism for salient object detection. Remote Sens 13(23):4941

    Google Scholar 

  28. Ker K et al (1982) Some negative results on the computational complexity of total variation and differentiation. Inf Control 53(1-2):21–31

    MathSciNet  Google Scholar 

  29. Kohout J (2007) On digital image representation by the delaunay triangulation. In: Pacific-rim symposium on image and video technology. Springer, pp 826–840

  30. Kreylos O, Hamann B (2001) On simulated annealing and the construction of linear spline approximations for scattered data. IEEE Trans Vis Comput Graphics 7(1):17–31

    Google Scholar 

  31. Lawonn K, Günther T (2019) Stylized Image Triangulation. In: Computer graphics forum. vol 38. Wiley Online Library, pp 221–234

  32. Li J, Yao L, Xu X, Cheng B, Ren J (2020) Deep reinforcement learning for pedestrian collision avoidance and human-machine cooperative driving. Information Sciences

  33. Liu X, Deng Z, Yang Y (2019) Recent progress in semantic image segmentation. Artif Intell Rev 52(2):1089–1106

    Google Scholar 

  34. Mark DB, Otfried C, Marc VK, Mark O (2008) Computational geometry algorithms and applications. Spinger

  35. Menandro FCM (2019) Two new classes of compactly supported radial basis functions for approximation of discrete and continuous data. Eng Rep 1(1):e12028

    Google Scholar 

  36. Mnih V, Kavukcuoglu K, Silver D, Graves A, Antonoglou I, Wierstra D et al (2013) Playing atari with deep reinforcement learning. arXiv:13125602

  37. Mnih V, Kavukcuoglu K, Silver D, Rusu AA, Veness J, Bellemare MG et al (2015) Human-level control through deep reinforcement learning. Nature 518(7540):529–533

    Google Scholar 

  38. Petrou M, Piroddi R, Talebpour A (2006) Texture recognition from sparsely and irregularly sampled data. Comput Vis Image Understand 102(1):95–104

    Google Scholar 

  39. Picard D, Revel A, Cord M (2012) An application of swarm intelligence to distributed image retrieval. Inf Sci 192:71–81

    Google Scholar 

  40. Prasad L, Skourikhine AN (2006) Vectorized image segmentation via trixel agglomeration. Pattern Recognit 39(4):501–514

    Google Scholar 

  41. Rajesh S, Sandeep K, Mittal R (2007) A fast progressive image sampling using lifting scheme and non-uniform B-splines. In: 2007 IEEE International symposium on industrial electronics. IEEE, pp 1645–1650

  42. Redmon J, Divvala S, Girshick R, Farhadi A (2016) You only look once: Unified, real-time object detection. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 779–788

  43. Rila L (1998) mage coding using irregular subsampling and Delaunay triangulation. In: Proceedings SIBGRAPI’98. International symposium on computer graphics, image processing, and vision (Cat. No. 98EX237). IEEE, pp 167–173

  44. Robinson JA (1997) Efficient general-purpose image compression with binary tree predictive coding. IEEE Trans Image Process 6(4):601–608

    Google Scholar 

  45. Rudin LI, Osher S, Fatemi E (1992) Nonlinear total variation based noise removal algorithms. Phys D: Nonlinear Phenom 60(1-4):259–268

    MathSciNet  Google Scholar 

  46. Sarkis M, Diepold K (2009) Content adaptive mesh representation of images using binary space partitions. IEEE Trans Image Process 18(5):1069–1079

    MathSciNet  Google Scholar 

  47. Scholefield A, Dragotti PL (2014) Quadtree structured image approximation for denoising and interpolation. IEEE Trans Image Process 23(3):1226–1239

    MathSciNet  Google Scholar 

  48. Silver D, Schrittwieser J, Simonyan K, Antonoglou I, Huang A, Guez A et al (2017) Mastering the game of go without human knowledge. Nature 550(7676):354–359

    Google Scholar 

  49. Skala V (2011) Incremental radial basis function computation for neural networks. WSEAS Trans Comput 10(11):367–378

    Google Scholar 

  50. Srinivasu PN, Balas VE (2021) Self-learning Network-based segmentation for real-time brain MR images through HARIS. PeerJ Comput Sci 7:e654

    Google Scholar 

  51. Srinivasu PN, Bhoi AK, Jhaveri RH, Reddy GT, Bilal M, Probabilistic deep Q (2021) Network for real-time path planning in censorious robotic procedures using force sensors. J Real-Time Image Process 18(5):1773–1785

    Google Scholar 

  52. Su D, Willis P (2004) Image interpolation by pixel-level data-dependent triangulation. In: Computer graphics forum. vol 23. Wiley Online Library, pp 189–201

  53. Sutton RS, Barto AG (2018) Reinforcement learning: An introduction. MIT press

  54. Tamang J, Nkapkop JDD, Ijaz MF, Prasad PK, Tsafack N, Saha A et al (2021) Dynamical properties of ion-acoustic waves in space plasma and its application to image encryption. IEEE Access 9:18762–18782

    Google Scholar 

  55. Viola P, Jones M (2001) Rapid object detection using a boosted cascade of simple features. In: Proceedings of the 2001 IEEE computer society conference on computer vision and pattern recognition. CVPR 2001. IEEE, vol 1

  56. Wang Z, Bovik AC, Sheikh HR, Simoncelli EP (2004) Image quality assessment: From error visibility to structural similarity. IEEE Trans Image Process 13(4):600–612

    Google Scholar 

  57. Wang K, Lo CP, Brook GA, Arabnia HR (2001) Comparison of existing triangulation methods for regularly and irregularly spaced height fields. Int J Geographical Inf Sci 15(8):743–762

    Google Scholar 

  58. Wang K, Zhao J, Feng J, Zhou B (2017) A three-dimensional error-diffusion algorithm for importance sampling with blue-noise property. In: International conference on computer graphics theory and applications. vol 2. SCITEPRESS, pp 70–81

  59. Wu W, Rui Y, Su F, Cheng L, Wang J (2014) Novel parallel algorithm for constructing Delaunay triangulation based on a twofold-divide-and-conquer scheme. GIScience Remote Sens 51(5):537–554

    Google Scholar 

  60. Wu Y, Wo Y, Han G (2022) Joint manipulation trace attention network and adaptive fusion mechanism for image splicing forgery localization. Multimed Tools Appl, pp 1–24

  61. Yang Y, Wernick MN, Brankov JG (2003) A fast approach for accurate content-adaptive mesh generation. IEEE Trans Image Process 12 (8):866–881

    MathSciNet  Google Scholar 

  62. Ye Z, Yi R, Gong W, He Y, Liu YJ (2020) Dirichlet energy of Delaunay meshes and intrinsic Delaunay triangulations. Comput-Aided Design 126:102851

    MathSciNet  Google Scholar 

  63. Yu X, Bryan B, Sederberg TW (2001) Image reconstruction using data-dependent triangulation. IEEE Comput Graphics Appl 21(3):62–68

    Google Scholar 

  64. Zapletal J, Vaněček P, Skala V (2009) RBF-based image restoration utilising auxiliary points. In: Proceedings of the 2009 Computer graphics international conference, pp 39–43

  65. da Silva ES, Santos A, Pedrini H (2018) Metrics for image surface approximation based on triangular meshes. Image Anal Stereology 37(1):71–82

    Google Scholar 

Download references

Acknowledgments

This work was supported by the Natural Science Foundation (NSF) of China (No. 61802147), and the Guangdong Basic and Applied Basic Research Foundation (No. 2022A1515011538 and 2022A1515012122). Authors would like to thank the anonymous referees for their useful comments, which were of great help in improving the exposition and readability of this paper.

Author information

Authors and Affiliations

  1. Department of Mathematics, Jinan University, Guangzhou, China

    Wenyong Gong & Xu-Qian Fan

Authors
  1. Wenyong Gong
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Xu-Qian Fan
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Xu-Qian Fan.

Ethics declarations

Conflict of Interests

The authors declared that they have no conflicts of interest to this work.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendix: A

Appendix: A

We repeatedly perform the proposed algorithm five times in the lena image (Fig. 4 (a)), pepers image (Fig. 4 (b)) and barbara image (Fig. 4 (c)), respectively. In the appendix, the repeated random experimental data is given in Tables 6, 7, 8.

Table 6 Five repeated experimental data in the lena image
Full size table
Table 7 Five repeated experimental data in the pepers image
Full size table
Table 8 Five repeated experimental data in the barbara image
Full size table

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Gong, W., Fan, XQ. Image adaptive sampling using reinforcement learning. Multimed Tools Appl 83, 5511–5530 (2024). https://doi.org/10.1007/s11042-023-15558-9

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11042-023-15558-9

Keywords

Search

Navigation