A Hybrid Approach to Local Contrast Enhancement Using Adaptive Neural Network Parameter Control

Relevance. Modern image processing techniques are focused on enhancing visual quality, particularly through adaptive local contrast enhancement. Previously, classical algorithms were employed to achieve high contrast efficiency; however, these approaches failed to account for the global scene context and often led to noise amplification. This paper proposes a hybrid method for adaptive local image contrast enhancement utilizing neural network-based parameter adjustment.

The aim of this research is to develop an algorithm that provides optimal contrast enhancement while minimizing noise artifacts and distortions, thereby improving contrast and real-time object detection accuracy.

The essence of the proposed solution lies in employing a convolutional neural network for automatic configuration of local contrast parameters based on statistical brightness characteristics and textural image features. The proposed method incorporates image segmentation into local regions, analysis of their properties, and adaptive adjustment of processing parameters. This results in improved discernibility of low-contrast objects under various imaging conditions. The algorithm's operating principle is based on dynamically selecting local region dimensions and contrast parameters depending on background and target scene objects. The integration of a neural network module enables precise adjustment of processing parameters while minimizing undesirable artifacts such as halos and blockiness. The methodology has been implemented as software and hardware for an optoelectronic system designed for computer vision applications, aerial image processing, video surveillance systems, and locating victims in various disaster scenarios.

The scientific novelty of this work lies in the development of an algorithm that automatically regulates contrast parameters based on analysis of both global and local scene context using artificial intelligence.

The theoretical significance of the work consists in the development of a contrast enhancement algorithm and image quality assessment method that accounts for contrast perception characteristics by both humans and AI systems under challenging observational conditions, such as fog, smoke, low illumination, etc.

The practical significance of the developed algorithm is determined by its implementation of contrast enhancement for objects in images acquired in both visible and infrared spectral ranges, and by the reliability of their recognition using artificial intelligence.

1. Jobson D.J., Rahman Z., Woodell G.A. Properties and performance of a center/surround retinex. IEEE Transactions on Image Processing. 1997;6(3):451‒462. DOI:10.1109/83.557356

2. Chen Y.S., Wang Y.C., Kao M.H., Chuang Y.Y. Deep Photo Enhancer: Unpaired Learning for Image Enhancement from Photographs with GANs. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, CVPR, 18‒23 June 2018, Salt Lake City, USA. IEEE; 2018. DOI:10.1109/CVPR.2018.00660

3. Paris S., Hasinoff S.W., Kautz J. Local Laplacian filtering: edge-aware image processing with a Laplacian pyramid. Proceedings of the Conference on Special Interest Group on Computer Graphics and Interactive Techniques, SIGGRAPH '11, 7‒11 August 2011, Vancouver, Canada. New York: Association for Computing Machinery; 2011. URL: https://people.csail.mit.edu/sparis/publi/2011/siggraph/Paris_11_Local_Laplacian_Filters_lowres.pdf [Accessed 25.04.2025]

4. Gritskevich I., Gogol A. No-Reference Image Quality Assessment Algorithm. Proceedings of Telecommunication Universities. 2024;10(2):16‒23. (in Russ.) DOI:10.31854/1813-324X-2024-10-2-16-23. EDN:TTPABW

5. Rec. ITU-R BT.500-11. Methodology for subjective assessment of the quality of television pictures. ITU-T. 2002.

6. Shelepin Yu.E. Introduction to Neuroiconics. St. Petersburg: Troickij most Publ.; 2017. 352 p. (in Russ.) EDN:YNTJRJ

7. Kim Y.T. Contrast enhancement using brightness preserving bi-histogram equalization. IEEE Transactions on Consumer Electronics. 1997;43:1. DOI:10.1109/30.580378

8. Rahman Z., Jobson D.J., Woodell G.A. Multi-scale retinex for color image enhancement. Proceedings of 3rd IEEE International Conference on Image Processing, 19 September 1996, Lausanne, Switzerland. IEEE; 1996. DOI:10.1109/ICIP.1996.560995

9. Ying Z., Li G., Gao W. A Bio-Inspired Multi-Exposure Fusion Framework for Low-light Image Enhancement. arXiv pre-print arXiv:1711.00591. 2017. DOI:10.48550/arXiv.1711.00591

10. Vala H.J., Baxi A. A review on Otsu image segmentation algorithm. International Journal of Advanced Research in Computer Engineering & Technology. 2013;2(2):387‒389.

11. Cybenko G. Approximation by superpositions of a sigmoidal function. Mathematics of Control, Signals and Systems. 1989;2:303–314. DOI:10.1007/bf02551274. EDN:OKSIPR

12. Wang R., Zhang Q., Fu C.W., Shen X., Zheng W.S., Jia J. Underexposed Photo Enhancement Using Deep Illumination Estimation. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR, 15‒20 June 2019, Long Beach, USA. IEEE; 2019. DOI:10.1109/CVPR.2019.00701

13. Han Y., Ye J.C. Framing U-Net via Deep Convolutional Framelets: Application to Sparse-View CT. IEEE Transactions on Medical Imaging. 2018;37(6):1418‒1429. DOI:10.1109/TMI.2018.2823768

14. Great Learning. What is Histogram Equalization and how it works? 2024. URL: https://www.mygreatlearning.com/blog/histogram-equalization-explained [Accessed 25.04.2025]

15. Liu J., Li D., Yuan C., Luo B., Wu G. A low-light image enhancement method with brightness balance and detail preservation. PLoS One. 2022;17(5):e0262478. DOI:10.1371/journal.pone.0262478. EDN:DFDSOY

16. Pizer S.M., Amburn E.P., Austin J.D., Cromartie R., Geselowitz A., Greer T., et al. Adaptive histogram equalization and its variations. Computer Vision, Graphics, and Image Processing. 1987;39(3):355‒368 DOI:10.1016/S0734-189X(87)80186-X

17. Kaur M., Kaur J., Kaur J. Survey of Contrast Enhancement Techniques based on Histogram Equalization. International Journal of Advanced Computer Science and Applications. 2011;2(7). DOI:10.14569/IJACSA.2011.020721

18. Zuiderveld K. VIII.5. ‒ Contrast Limited Adaptive Histogram Equalization. In: Heckbert P.S. (ed.) Graphics Gems IV. Academic Press; 1994. p.474‒485. DOI:10.1016/B978-0-12-336156-1.50061-6

19. Lore K.G., Akintayo A., Sarkar S. LLNet: A Deep Autoencoder Approach to Natural Low-light Image Enhancement. Pattern Recognition. 2017;61:650‒662. DOI:10.1016/j.patcog.2016.06.008

20. Wei C., Wang W., Yang W., Liu J. Deep Retinex Decomposition for Low-Light Enhancement. 2018. URL: http://39.96.165.147/Pub%20Files/2018/chen_bmvc18.pdf [Accessed 25.04.2025]

21. Liu X., Ma Y., Shi Z., Chen J. GridDehazeNet: Attention-Based Multi-Scale Network for Image Dehazing. Proceedings of the IEEE/CVF International Conference on Computer Vision, ICCV, 27 October ‒ 02 November 2019, Seoul, South Korea. IEEE; 2019. DOI:10.1109/ICCV.2019.00741

22. Guo C., Li C., Guo J., Loy C.C., Hou J., Kwong S., Cong R. Zero-Reference Deep Curve Estimation for Low-Light Image Enhancement. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR, 13‒19 June 2020, Seattle, USA. IEEE; 2020. DOI:10.1109/CVPR42600.2020.00185

23. Hossain F., Alsharif M.R. Image Enhancement Based on Logarithmic Transform Coefficient and Adaptive Histogram Equalization. Proceedings of the 2007 International Conference on Convergence Information Technology, ICCIT 2007, 21‒23 November 2007, Gwangju, South Korea. IEEE; 2007. DOI:10.1109/ICCIT.2007.4420457

24. Stark J.A. Adaptive image contrast enhancement using generalizations of histogram equalization. IEEE Transactions on Image Processing. 2000;9(5):889‒896. DOI:10.1109/83.841534

25. Potapova A.A. The Latest Methods of Image Processing. Moscow: FIZMATLIT Publ.; 2008. 496 p. (in Russ.)