In sandstorm weather, outdoor images often appear blurry, with color cast, etc., which seriously affects the performance of outdoor computer vision systems. The current sand and dust image enhancement algorithms often suffer from color distortion in large areas of the sky. A multi-scale information and HSV (Hue, Saturation, Value) space-based image denoising algorithm was designed for this purpose. The multi-scale residual denoising module was used to remove sand and dust, and the HSV global adjustment module was used to further adjust the image color cast. Then, an attention mechanism-based feature fusion module was used to fuse the outputs of the two modules according to their corresponding weights, restoring high-quality images. The experimental results show that the proposed sand and dust image enhancement algorithm eliminates color cast in the image and achieves good results in both subjective and objective evaluations.
NIUHongxia, WANGChunzhi, LIANGLeguan, et al. Sand dust image clarity algorithm based on improved dark channel prior[J]. Journal of Jilin University (Science Edition), 2023, 61(6): 1407-1418. (in Chinese)
[3]
XIANGP, CHENC, LIUG, et al. Image enhancement of degraded sand-dust images based on channel compensation and brightness partitioning[C]//2023 2nd International Conference on Robotics, Artificial Intelligence and Intelligent Control (RAIIC), 2023: 288-293.
WANGHe, HANLei. Image dehazing algorithm based on Res2net and pyramid pooling[J]. Journal of Test and Measurement Technology, 2023, 37(5): 455-460. (in Chinese)
CHENWanqiang, LUYonghua, ZHUYun, et al. Research on surface defect detection algorithm of fuel pump regulator based on deep learning image registration[J].Journal of Test and Measurement Technology, 2024, 38(6): 586-592. (in Chinese)
[8]
ZHANGR, CAOZ, YANGS, et al. Cognition-driven structural prior for instance-dependent label transition matrix estimation[J]. IEEE Transactions on Neural Networks and Learning Systems, 2025, 36(2): 3730-3743.
[9]
DINGB, ZHANGR, XUL, et al. U2D2Net: unsupervised unified image dehazing and denoising network for single hazy image enhancement[J]. IEEE Transactions on Multimedia, 2023, 26: 202-217.
[10]
LIANGP, DONGP, WANGF, et al. Learning to remove sandstorm for image enhancement[J]. The Visual Computer, 2023, 39(5): 1829-1852 .
[11]
DINGY, WUK. USIR-Net: sand-dust image restoration based on unsupervised learning[J]. Machine Vision and Applications, 2024, 35(3): 45.
SHIZhenghao, LIUChunyue, RENWenqi, et al. Convolutional neural networks for sand dust image color restoration and visibility enhancement[J]. Journal of Image and Graphics, 2022, 27(5): 1493-1508. (in Chinese)
DINGBosheng, ZHANGRuiheng, XULixin, et al. Sand-dust image restoration using gray compensation and feature fusion[J]. Acta Armamentarii, 2023, 44(10): 3115-3126. (in Chinese)
[18]
SIY, YANGF, GUOY, et al. A comprehensive benchmark analysis for sand dust image reconstruction[J]. Journal of Visual Communication and Image Representation, 2022, 89: 103638.
[19]
BHATS F, BIRKLR, WOFKD, et al. ZoeDepth: zero-shot transfer by combining relative and metric depth[EB/OL]. Computer Science, 2023. arXiv: 2302.12288.
LICe, LIUHao, CHENXiaolei, et al. Enhancement algorithm of dust image based on multi-perceptual feature calculation[J]. Journal of Lanzhou University of Technology, 2018, 44(4): 90-95. (in Chinese)
[22]
SHIZ, FENGY, ZHAOM, et al. Let you see in sand dust weather: a method based on halo-reduced dark channel prior dehazing for sand-dust image enhancement[J]. IEEE Access, 2019, 7: 116722-116733.
[23]
KANTI DHARAS, ROYM, SEN D, et al. Color cast dependent image dehazing via adaptive airlight refinement and non-linear color balancing[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2021, 31(5): 2076-2081.
[24]
LIUJ, LIUR W, SUNJ, et al. Rank-one prior: real-time scene recovery[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023, 45(7): 8845-8860.
[25]
QUY, CHENY, HUANGJ, et al. Enhanced Pix2pix dehazing network[C]//2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019: 8152-8160.
[26]
ZHAOS, ZHANGL, SHENY, et al. RefineDNet: a weakly supervised refinement framework for single image dehazing[J]. IEEE Transactions on Image Processing, 2021, 30: 3391-3404.
[27]
GAOY, XUW, LUY. Let you see in haze and sandstorm: two-in-one low-visibility enhancement network[J]. IEEE Transactions on Instrumentation and Measurement, 2023, 72: 5023712.
[28]
LUY, YANGD, GAOY, et al. AoSRNet: all-in-one scene recovery networks via multi-knowledge integration[EB/OL]. Computer Science, 2024. arXiv: 2402.03738.