Research on High-Resolution Image Harmonization Method Based on Multi-Scale and Global Feature Guidance
Source abstract
Abstract
During the image compositing process, there may be inconsistencies in tone and illumination between the foreground and background, leading to poor visual quality and low realism in the composite images. To address these issues, image harmonization techniques can be employed. This paper proposes an image harmonization method based on multi-scale and global feature guidance (MSGF). In general, images captured in different scenes may exhibit inconsistencies in lighting after composition. The goal of image harmonization is to adjust the foreground illumination to match that of the background. Traditional methods often attempt to blend pixels directly, which can result in unrealistic outcomes. The proposed approach combines multi-scale feature extraction with global feature guidance, forming the MSGF framework. The experiment was conducted on the iHarmony4 dataset. Comparative experiments showed that MSGF achieved the best performance on three subset indicators, including HCOCO. Ablation studies demonstrated the effectiveness of the proposed module. Efficiency evaluation results indicated that it took 0.01s and had 20.9 million parameters, outperforming comparative methods and effectively achieving high-quality image harmonization.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Advanced Image Processing Techniques · Advanced Image Fusion Techniques · Image Processing Techniques and Applications
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
School of Computer Science, Qinghai Normal University, Xining 810000, China
School of Computer and Software, Nanyang Institute of Technology, Nanyang 473000, China
Academy of Plateau Science and Sustainability, People’s Government of Oinghai Province & Beijing Normal University, Haihu, Xining 810004, China
The State Key Laboratory of Tibetan Intelligent Information Processing and Application, Qinghai Normal University, Hutai, Xining 810008, China
Funding
Natural Science Foundation for Youth of Qinghai Province · National Natural Science Foundation of China
References (23)
Show source-provided references
1. Multi-scale image harmonization10.1145/1778765.1778862 ↗
2. Deep painterly harmonization10.1111/cgf.13478 ↗
3. Understanding and improving the realism of image composites10.1145/2185520.2185580 ↗
4. Sofiiuk, K., Popenova, P., and Konushin, A. (2021, January 5–9). Foreground-aware semantic representations for image harmonization. Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, Virtual.10.1109/WACV48630.2021.00166 ↗
5. Ke, Z., Sun, C., Zhu, L., Xu, K., and Lau, R.W.H. (2022, January 23–27). Harmonizer: Learning to perform white-box image and video harmonization. Proceedings of the European Conference on Computer Vision, Tel Aviv, Israel.10.1007/978-3-031-19784-0_40 ↗
6. Cong, W., Zhang, J., Niu, L., Liu, L., Ling, Z., Li, W., and Zhang, L. (2020, January 14–19). Dovenet: Deep image harmonization via domain verification. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, WA, USA.10.1109/CVPR42600.2020.00842 ↗
7. Guo, Z., Guo, D., Zheng, H., Gu, Z., Zheng, B., and Dong, J. (2021, January 11–17). Image harmonization with transformer. Proceedings of the IEEE/CVF International Conference on Computer Vision, Virtual.10.1109/ICCV48922.2021.01460 ↗
8. Wang, K., Gharbi, M., Zhang, H., Xia, Z., and Shechtman, E. (2023, January 18–22). Semi-supervised parametric real-world image harmonization. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Vancouver, BC, Canada.10.1109/CVPR52729.2023.00574 ↗
9. High-resolution image harmonization with adaptive-interval color transformation
10. Poisson image editing
11. Tian, C., and Zhang, Q. (2025, January 6–11). Self-Supervised Image Harmonization via Holistic Feature Fusion. Proceedings of the ICASSP 2025—2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Hyderabad, India.10.1109/ICASSP49660.2025.10889606 ↗
12. Usama, M., Nyman, E., Näslund, U., and Grönlund, C. (2025). A domain adaptation model for carotid ultrasound: Image harmonization, noise reduction, and impact on cardiovascular risk markers. Comput. Biol. Med., 190.10.1016/j.compbiomed.2025.110030 ↗
13. Vision transformer-based feature harmonization network for fine-resolution land cover mapping10.1080/01431161.2025.2491816 ↗
14. Harmonize: A shared environment for extended immersive entertainment10.1007/s10055-021-00585-4 ↗
15. Duan, L., Wu, M., Lou, H., Yin, J., and Li, X. (2024, January 6–10). MRCAN: Multi-scale Region Correlation-driven Adaptive Normalization for Image Harmonization. Proceedings of the 2024 IEEE International Conference on Systems, Man, and Cybernetics (SMC), Kuching, Malaysia.10.1109/SMC54092.2024.10831083 ↗
16. Cong, W., Tao, X., Niu, L., Liang, J., Gao, X., Sun, Q., and Zhang, L. (2022, January 19–24). High-resolution image harmonization via collaborative dual transformations. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, LA, USA.10.1109/CVPR52688.2022.01792 ↗
17. Niu, L., Tan, L., Tao, X., Cao, J., Guo, F., Long, T., and Zhang, L. (2023, January 2–3). Deep image harmonization with globally guided feature transformation and relation distillation. Proceedings of the IEEE/CVF International Conference on Computer Vision, Paris, France.10.1109/ICCV51070.2023.00710 ↗
18. Yuan, J., Wu, H., Xie, L., Zhang, L., and Xing, J. (2025). Learning multi-color curve for image harmonization. Eng. Appl. Artif. Intell., 146.10.1016/j.engappai.2025.110277 ↗
19. Cong, W., Zhang, J., Niu, L., Liu, L., Ling, Z., Li, W., and Zhang, L. (2019). Image Harmonization Dataset iHarmony4: HCOCO, HAdobe5k, HFlickr, and Hday2night. arXiv.
20. Lin, T.Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Dollár, P., and Zitnick, C.L. (2014, January 6–12). Microsoft coco: Common objects in context. Proceedings of the Computer Vision—ECCV 2014: 13th European Conference, Zurich, Switzerland. Part v 13.10.1007/978-3-319-10602-1_48 ↗
21. Semantic understanding of scenes through the ade20k dataset10.1007/s11263-018-1140-0 ↗
22. Bychkovsky, V., Paris, S., Chan, E., and Durand, F. (2011, January 20–25). Learning photographic global tonal adjustment with a database of input/output image pairs. Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2011), Colorado Springs, CO, USA.10.1109/CVPR.2011.5995332 ↗
23. Tsai, Y.-H., Shen, X., Lin, Z., Sunkavalli, K., Lu, X., and Yang, M.-H. (2017, January 21–26). Deep image harmonization. Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2017), Honolulu, HI, USA.10.1109/CVPR.2017.299 ↗