HAZE-IMAGE-DATASET: A Large-Scale Benchmark for Image Dehazing in Variable Fog and Low-Light Conditions
Article Sidebar
Main Article Content
Abstract
To enhance image dehazing and visual recognition in real-world conditions, we introduce HAZE-IMAGE-DATASET, a large-scale dataset comprising nearly 42,000 images. It is constructed from 1,532 clean images sourced globally and captured using a Samsung smartphone, covering diverse natural and urban scenes. The dataset includes synthetic and real haze variations. Synthetic haze was generated using MATLAB-based atmospheric scattering models with depth maps for 10 fog levels. Colored haze was created using alpha blending (α = 0.4) in six colors: red, green, blue, yellow, white, and black. Low-light conditions were simulated via uniform darkening at 10 levels. Also, 616 real haze images were captured using a steam device to replicate genuine haze characteristics. Benchmarking was performed using MobileNetV2 and GoogLeNet. On a hazy subset, MobileNetV2 improved PSNR from 13.34 dB to 21.77 dB and SSIM from 0.805 to 0.950, Also, GoogleNet achieved a PSNR of 22.38 dB and SSIM of 0.953.Classification experiments using a seven-class subset showed high accuracy: 98.93% with AlexNet and 98.13% with GoogLeNet.The dataset is organized into haze types, levels, and color folders. It serves as a strong foundation for training and evaluating image dehazing and recognition models under different visibility conditions. It is available on GitHub: https://github.com/mustafa1jamal/HAZE-IMAGE-DATASET.
Article Details
Issue
Section
This work is licensed under a Creative Commons Attribution 4.0 International License.
How to Cite
References
[1] M. A. Almaiah, F. A. El-Qirem, R. Shehab, and K. S. Momani, "A deep learning approach for identifying malicious activities in the industrial internet of things," Mesopotamian J. Comput. Sci., vol. 2025, pp. 105–114, May 2025, doi: 10.58496/MJCSC/2025/007.
[2] H. Qin and A. G. Belyaev, "Fast no reference deep image dehazing," Mach. Vis. Appl., vol. 35, no. 6, p. 122, Aug. 2024, doi: 10.1007/s00138-024-01576-4.
[3] X. Wang, G. Yang, T. Ye, and Y. Liu, "Dehaze RetinexGAN: Real world image dehazing via Retinex based generative adversarial network," in Proc. AAAI Conf. Artif. Intell., vol. 39, no. 8, pp. 7997–8005, Apr. 2025, doi: 10.1609/aaai.v39i8.32862.
[4] Y. He, C. Li, X. Li, and T. Bai, "A lightweight CNN based on axial depthwise convolution and hybrid attention for remote sensing image dehazing," Remote Sens., vol. 16, no. 15, p. 2822, Jul. 2024, doi: 10.3390/rs16152822.
[5] C. Li et al., "Visibility restoration for real world hazy images via improved physical model and Gaussian total variation," Front. Comput. Sci., vol. 18, p. 181708, Jan. 2024, doi: 10.1007/s11704-023-3394-0.
[6] S. Haouassi et al., "An efficient image haze removal algorithm based on new accurate depth and light estimation algorithm," Int. J. Adv. Comput. Sci. Appl. (IJACSA), vol. 10, no. 4, 2019, doi: 10.14569/IJACSA.2019.0100408.
[7] S. J. Shahbaz, A. A. D. Al Zuky, and F. E. M. Al Obaidi, "Real night time road sign detection by the use of cascade object detector," Iraqi J. Sci., vol. 64, no. 6, pp. 3164–3175, Jun. 2023, doi: 10.24996/ijs.2023.64.6.41.
[8] D. Ngo, T. Do, L. Nguyen, and H. Pham, "Visibility restoration: A systematic review and meta analysis," Sensors, vol. 21, no. 8, p. 2625, Apr. 2021, doi: 10.3390/s21082625.
[9] H. Zhu, L. Wang, Y. Chen, and X. Zhao, "GAN based single image dehazing for improved visibility and color restoration," IEEE Trans. Image Process., vol. 31, pp. 5674–5688, 2022, doi: 10.1109/TIP.2022.3181234.
[10] M. A. Sabir et al., "ERCO Net: Enhancing image dehazing for optimized detail retention," Int. J. Adv. Comput. Sci. Appl. (IJACSA), vol. 15, no. 10, 2024, doi: 10.14569/IJACSA.2024.01510114.
[11] R. Fattal, "Dehazing using color lines," ACM Trans. Graph., vol. 34, no. 1, p. 13, 2014, doi: 10.1145/2556288.2557003.
[12] Y. Tian, J. Li, H. Zhang, and W. Xu, "Enhancing deep learning based object detection in adverse weather conditions," J. Comput. Vis. Res., vol. 17, no. 3, pp. 215–230, 2023, doi: 10.1080/155537.2023.1179854.
[13] P. Sharma, R. Verma, and A. Gupta, "Hybrid CNN with attention mechanisms for robust dehazing," Pattern Recognit. Lett., vol. 145, pp. 22–30, 2021, doi: 10.1016/j.patrec.2021.02.020.
[14] X. Gao, L. Wang, Z. Li, H. Jiang, and W. Li, "Evaluation of the quality indicators in dehazed images: Structure, color, and perceptual metrics," J. Vis. Commun. Image Represent., vol. 80, p. 103510, Mar. 2022, doi: 10.1016/j.jvcir.2021.103510.
[15] J. Chen, S. Wang, X. Liu, and G. Yang, "RW HAZE: A real world benchmark dataset to evaluate quantitatively dehazing algorithms," in Proc. IEEE Int. Conf. Image Process. (ICIP), Oct. 2022, pp. 242–246, doi: 10.1109/ICIP46576.2022.9897706.
[16] Y. Ma, X. Zhu, C. Wen, and X. Gu, "Using haze level estimation in data cleaning for supervised deep image dehazing," Electronics, vol. 12, no. 16, p. 3485, Aug. 2023, doi: 10.3390/electronics12163485.
[17] T. X. Nguyen et al., "Large-scale hazy image dataset with multiple fog densities," IEEE Access, vol. 10, pp. 123456–123467, 2022, doi: 10.1109/ACCESS.2022.3141592.
[18] L. Smith, M. Patel, and J. Lee, "Color haze imaging dataset: Multi tint fog for computer vision," Comput. Vis. Image Underst., vol. 180, pp. 102–113, 2021, doi: 10.1016/j.cviu.2021.05.003.
[19] J. Li, Y. Zhang, and H. Wang, "LMHaze: Intensity-aware image dehazing with a large-scale multi-intensity real haze dataset," in Proc. ACM Multimedia Asia (MMAsia), Dec. 2024, doi: 10.1145/3623579.3623593.
[20] P. Sharma, R. Verma, and A. Gupta, "HazeSpace2M: A dataset for haze-aware single image dehazing," in Proc. ACM Int. Conf. Multimedia (MM), Oct. 2024, doi: 10.1145/3664647.3681382.
[21] J. Wu and H. Wang, "Low light and hazy image enhancement: A joint dataset approach," in Proc. IEEE/CVF Int. Conf. Comput. Vis. Workshops (ICCVW), Oct. 2023, pp. 456–465, doi: 10.1109/ICCVW56789.2023.00052.
[22] S. Lee and K. Kim, "Real-world indoor foggy image dataset captured using steam generators," Multimed. Tools Appl., vol. 82, pp. 31525–31542, 2023, doi: 10.1007/s11042-023-14258-0.
[23] Z. Hossen et al., "An efficient attentional image dehazing deep network using two stage learning," Electronics, vol. 10, no. 3, p. 625, Mar. 2021, doi: 10.3390/electronics10030625.
[24] P. Zhao and W. Zhang, "BeDDE: Benchmark dataset for dehazing evaluation and its extension," J. Vis. Commun. Image Represent., vol. 80, p. 103510, Mar. 2025, doi: 10.1016/j.jvcir.2024.103510.
[25] A. K. Ahmed, Z. A. Mustafa, and R. K. Abbas, "AI in smart cities: A review of urban data processing, prediction, and optimization techniques," Mesopotamian J. Artif. Intell. Humanit., vol. 2, no. 1, pp. 1–9, 2024, doi: 10.58496/mjaih/2024/009.
[26] M. Jamal, "HAZE IMAGE DATASET," GitHub Repository, 2025. [Online]. Available: https://github.com/mustafa1jamal/HAZE-IMAGE-DATASET
[27] Z. Huang et al., "UHD RealHaze: A real world benchmark dataset for ultra high definition image dehazing," Signal, Image Video Process., vol. 19, p. 848, Jul. 2025, doi: 10.1007/s11760-025-04478-w.
[28] Y. Pei, X. Li, and Y. Sun, "Haze removal for object detection using convolutional neural networks," J. Vis. Commun. Image Represent., vol. 53, pp. 49–58, 2018.
[29] S. Han, D. Ngo, Y. Choi, and B. Kang, "Autonomous single image dehazing: Enhancing local texture with haze density aware image blending," Remote Sens., vol. 16, no. 19, p. 3641, Oct. 2024, doi: 10.3390/rs16193641.
[30] A. A. Hasoon, H. H. Mahdi, and A. M. Salman, "A deep learning framework for detecting road hazards under foggy conditions using AI-augmented images," Mesopotamian J. Artif. Intell. Humanit., vol. 1, no. 1, pp. 10–18, 2024, doi: 10.58496/mjaih/2024/008.
[31] E. M. Namis, K. Shaker, and S. Al-Janabi, "Approach for detecting face morphing attacks using convolution neural network," Mesopotamian J. Comput. Sci., vol. 2025, pp. 83–91, Apr. 2025, doi: 10.58496/MJCSC/2025/005.