Removing non-homogeneous haze from a single image without losing structural and color fidelity is a major challenge due to the lack of natural scene priors. To address this problem, we propose a novel Gated Graph Reasoning Network with Color Tuning and Calibration (GCTC-Net). Specifically, to eliminate the information redundancy of traditional rigid feature stacking, an Adaptive Gated Attention Interaction (AGAI) module is introduced to adaptively aggregate multi-exposure Gated Contrast-enhanced Prior (GCP) maps with primary hazy features via pixel-level attention. Subsequently, a dual-path topo-logical graph reasoning backbone is deployed to model non-local dependencies across spatial and channel dimensions, effectively propagating clean structural clues to heavily contaminated nodes. Finally, a dedicated non-linear Color Calibration Module (CCM) is incorporated at the back end to execute localized color variance restoration, successfully suppressing cumulative chromatic deviations and white-balance shifts. Extensive experiments on public benchmarks demonstrate that our proposed GCTC-Net achieves superior performance and establishes an optimal balance between structural sharpness and realistic color fidelity.
References
[1] Yuan, H., Yan, T. H. (2024) Parallel multi scale attention mapping image dehazing algorithm. Laser & Optoelectronics Progress, 61(4), 101-112.
[2] Wang, Q., Jiang, K., Wang, Z., Ren, W., Zhang, J., Lin, C. W. (2023) Multi-scale fusion and decomposition network for single image deraining. IEEE Transactions on Image Processing, 33, 191-204.
[3] Minaee, S., Boykov, Y., Porikli, F., Plaza, A., Kehtarnavaz, N., Terzopoulos, D. (2021) Image segmentation using deep learning: a survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(7), 3523-3542.
[4] Zhou, R., Wang, R., Wang, Y., Zhou, F., Guo, N. (2024) Research on low-light image enhancement based on MER-Retinex algorithm. Signal, Image and Video Processing, 18(1), 803-811.
[5] Othman, G., Zeebaree, D. Q. (2020) The applications of discrete wavelet transform in image processing: a review. Journal of Soft Computing and Data Mining, 1(2), 31-43.
[6] Lisani, J. L., Morel, J. M., Petro, A. B., Sbert, C. (2020) Analyzing center/surround retinex. Information Sciences, 512, 741-759.
[7] Singh, D., Kumar, V. (2018) Single image haze removal using integrated dark and bright channel prior. Modern Physics Letters B, 32(04), 1850051.
[8] Zhu, Q., Mai, J., Shao, L. (2015) A fast single image haze removal algorithm using color attenuation prior. IEEE Transactions on Image Processing, 24(11), 3522-3533.
[9] Huang, Z., Jing, H., Chen, A., Hong, C., Shang, X. (2023) Efficient image dehazing algorithm using multiple priors constraints. Journal of Visual Communication and Image Representation, 90, 103694.
[10] Yang, A., Liu, Y., Wang, J., Li, X., Cao, J., Ji, Z., Pang, Y. (2023) Visual-quality-driven unsupervised image dehazing. Neural Networks, 167, 1-9.
[11] Zhu, X., Ning, J., Wu, X., Duo, B. (2025) Frequency attention-based multi-scale dehazing GAN. Digital Signal Processing, 105747.
[12] Li, J., Zheng, M., Sun, C., Wang, Z., Shen, H. (2026) T-SFANet: A spatial-frequency attention network for thermal infrared image dehazing. Neurocomputing, 132797.
[13] Cai, B., Xu, X., Jia, K., Qing, C., Tao, D. (2016) Dehazenet: An end-to-end system for single image haze removal. IEEE Transactions on Image Processing, 25(11), 5187-5198.
[14] Chen, Z., He, Z., Lu, Z. M. (2024) DEA-Net: Single image dehazing based on detail-enhanced convolution and content-guided attention. IEEE Transactions on Image Processing, 33, 1002-1015.
[15] Batool, I., Imran, M. (2025) A dual residual dense network for image denoising. Engineering Applications of Artificial Intelligence, 147, 110275.
[16] Hu, B. (2021) Multi-scale feature fusion network with attention for single image dehazing. Pattern Recognition and Image Analysis, 31(4), 608-615.
[17] Park, J., Han, D. K., Ko, H. (2020) Fusion of heterogeneous adversarial networks for single image dehazing. IEEE Transactions on Image Processing, 29, 4721-4732.
[18] Zheng, L., Li, Y., Zhang, K., Luo, W. (2022) T-Net: Deep stacked scale-iteration network for image dehazing. IEEE Transactions on Multimedia, 25, 6794-6807.
[19] Zhang, X., Song, M., Fan, Z., Jin, H. (2022) Reconstruction algorithm of haze image based on blind separation model of polarized orthogonal airlight. Optics Express, 30(23), 42097-42113.
[20] Hou, Y., Xu, J., Liu, M., Liu, G., Liu, L., Zhu, F., Shao, L. (2020) NLH: A blind pixel-level non-local method for real-world image denoising. IEEE Transactions on Image Processing, 29, 5121-5135.
[21] Dong, P., Wang, B. (2022) TransRA: Transformer and residual attention fusion for single remote sensing image dehazing. Multidimensional Systems and Signal Processing, 33(4), 1119-1138.
[22] Li, B., Ren, W., Fu, D., Tao, D., Feng, D., Zeng, W., Wang, Z. (2018) Benchmarking single-image dehazing and beyond. IEEE Transactions on Image Processing, 28(1), 492-505.
[23] Sakaridis, C., Dai, D., Van Gool, L. (2018) Semantic foggy scene understanding with synthetic data. International Journal of Computer Vision, 126(9), 973-992.
[24] Nicolicioiu, A., Duta, I., Leordeanu, M. (2019) Recurrent space-time graph neural networks. Advances in Neural Information Processing Systems, 32.
[25] Chen, S., Li, Z., Tang, Z. (2020) Relation R-CNN: A graph based relation-aware network for object detection. IEEE Signal Processing Letters, 27, 1680-1684.
[26] Wang, R., Huang, C., Wang, X. (2020) Global relation reasoning Graph Convolutional Networks for human pose estimation. IEEE Access, 8, 38472-38480.
[27] Shafiq, M., Gu, Z. (2022) Deep residual learning for image recognition: a survey. Applied Sciences, 12(18), 8972.
[28] Dai, D., Sakaridis, C., Hecker, S., Van Gool, L. (2020) Curriculum model adaptation with synthetic and real data for semantic foggy scene understanding. International Journal of Computer Vision, 128(5), 1182-1204.
[29] Li, Y., Wei, X., Liao, X., Zhao, Y., Jia, F., Zhuang, X., Zhou, M. (2024) A deep retinex-based low-light enhancement network fusing rich intrinsic prior information. ACM Transactions on Multimedia Computing, Communications and Applications, 20(11), 1-23.
Share and Cite
Hu, J., Li, W., Liang, H. (2026) Gated Graph Reasoning Network with Color Tuning and Calibration for Image Dehazing. Scientific Research Bulletin, 3(2), 77-89. https://doi.org/10.71052/srb2024/QBJI3113